UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Large scale pullout testing of geosynthetics Muthu, Raju D. 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A7 M87.pdf [ 4.46MB ]
Metadata
JSON: 831-1.0062786.json
JSON-LD: 831-1.0062786-ld.json
RDF/XML (Pretty): 831-1.0062786-rdf.xml
RDF/JSON: 831-1.0062786-rdf.json
Turtle: 831-1.0062786-turtle.txt
N-Triples: 831-1.0062786-rdf-ntriples.txt
Original Record: 831-1.0062786-source.json
Full Text
831-1.0062786-fulltext.txt
Citation
831-1.0062786.ris

Full Text

L A R G E S C A L E P U L L O U T TESTING OF GEOSYNTHETICS By Raju D. Muthu B. E., Bangalore University, 1986 M . E., Indian Institute of Science, Bangalore, 1988 A THESIS S U B M I T T E D IN PARTIAL F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R S O F A P P L I E D S C I E N C E in T H E F A C U L T Y O F G R A D U A T E STUDIES Department of Civil Engineering We accept this thesis as conforming to the required standard T H E UNIVERSITY O F BRITISH C O L U M B I A September 1991 © Raju D. Muthu, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of C . | \ J l L Sf/^CfJL\AJ2JZK^J The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract An evaluation of soil-geosynthetic interface friction is important to the design of any anchorage detail of a reinforced soil structure or membrane-lined waste contain-ment facility. A large pullout apparatus has been designed and commissioned to evaluate the mobilization of pullout resistance in geosynthetic test specimens. Sand samples were prepared by pluviation into a rectangular box, 1.30m x 0.64m x 0.60m. A stress controlled top boundary was used to apply vertical stresses in the range 5 to 90 kPa. A rate of pullout displacement of 0.5 mm/min was used in the program of testing. A technique of strain gauging the geosynthetic test specimen has been developed. Variables examined in the program of testing were type of geosynthetic and con-fining stress. Measurement of pullout force, pullout displacement, horizontal pressure on front face of the test box, strain in geosynthetic material, water pressure in the surcharge bag, and volume change were taken during testing. Pullout resistance in-creases with confining stress and is described by a bond factor or a bond coefficient. Some test specimens failed in pullout, and some were tending toward tensile yield. A development of progressive strain was observed. n Table of Contents Abstract ii List of Tables vii List of Figures viii List of Symbols xi Acknowledgements xiii 1 Introduction 1 1.1 Designing with Geosynthetics 1 1.2 Research Needs 2 1.3 Thesis Organization 3 2 Literature Review 4 2.1 Introduction 4 2.2 Technique of Soil Reinforcement 4 2.3 Technique of Lining with Geosynthetics 6 2.4 Factors Influencing Pullout Resistance 9 2.4.1 Soil Characteristics 9 2.4.2 Geosynthetic Characteristics 10 2.4.3 Interaction between Soil and Geosynthetic in Pullout 11 2.5 Laboratory Studies 19 2.5.1 Pullout Tests 20 iii 2.5.1.1 Relative Density and Dilatancy Characteristics of the Soil 20 2.5.1.2 Confining Pressure 22 2.5.1.3 Grid Orientation 22 2.5.1.4 Geosynthetic Extensibility 23 2.5.1.5 Boundary Effects 23 2.6 Field Trials 24 2.7 Conclusions on Research Needs 27 3 Apparatus 28 3.1 Introduction 28 3.2 Large Pullout Apparatus . 28 3.2.1 Pullout box 29 3.2.2 Hopper 31 3.2.3 Pullout Load Assembly 33 3.2.3.1 Clamp 35 3.2.4 Surcharge Loading 35 3.2.4.1 Reaction 37 3.3 Instrumentation 37 3.3.1 Pullout Force 37 3.3.2 Pullout Displacement 39 3.3.3 Pressure on the Front Face 39 3.3.4 Water Pressure Transducer 39 3.3.5 Strain Gauges 40 3.3.6 Data Acquisition System 41 4 Material Properties 43 4.1 Introduction 43 iv 4.2 Properties of the Sand 43 4.3 Properties of the Geosynthetics 45 4.3.1 Geogrid Reinforcement 45 4.3.2 Geomembrane Liner 47 5 Experimental Procedure 50 5.1 Introduction and Test Program 50 5.2 Test Preparation 50 5.2.1 Laboratory Test Box 50 5.2.2 Geosynthetic Specimen 51 5.2.3 Sand Placement 51 5.2.4 Surcharge Loading 52 5.3 Pullout Assembly 53 5.4 Test Procedure 53 5.5 Post Test Routine 54 6 Results 55 6.1 Introduction 55 6.2 Results of Tests on the Geogrid 56 6.2.1 Density 56 6.2.2 Pullout Force 57 6.2.3 Local Strain 59 6.2.4 Passive Pressure on the Front Face 62 6.2.5 Volume Change 62 6.2.6 Repeatability 66 6.3 Results of Tests on the Geomembrane 68 6.3.1 Density 70 6.3.2 Pullout Force 70 v 6.3.3 Local Strain 72 6.3.4 Passive Pressure on the Front Face 75 6.3.5 Volume change . . 75 7 Analysis of Results 79 7.1 Introduction 79 7.2 Pullout Resistance of the Geogrid 79 7.3 Pullout Resistance of the Geomembrane 83 8 Conclusions 88 8.1 On Mobilization of Pullout Resistance 88 8.2 On Improvements to the Test Procedure 89 8.3 On Suggestions for Future Work 89 References 91 Appendices 96 A Technique of Strain Gauging Plastics 96 A.1 Introduction 96 A.2 Characteristics of the Strain Gauge 97 A.3 Strain Gauging Procedure 98 A.4 Chemicals for Surface Preparation 98 A.5 Adhesive Selection 98 A.5.1 Geosynthetic Surface Preparation 98 A.5.2 Gauge Preparation 99 A.5.3 Application of the Gauge 100 A.5.4 Gauge Soldering 101 A.5.5 Gauge Protection 102 A.6 Analysis of Strain Data 102 vi List of Tables 2.1 Summary of pullout Tests (modified from Juran et al., 1988) 25 4.1 Properties of the Tensar UX1500 Geogrid 46 4.2 Properties of the Novex Smooth Surface H D P E Geomembrane . . . . 49 6.1 Geogrid Test Parameters 56 6.2 Geomembrane Test Parameters 69 A . l Dimensions of the Strain Gauge 97 vii List of Figures 2.1 Typical Examples of Soil Reinforcement Applications (after Palmeira, 1987) 5 2.2 Failure Mechanisms in a Reinforced Soil Retaining Wall (after Palmeira, 1987) 6 2.3 Section Through a Liner System (after Udwari and Kittridge, 1986) . 7 2.4 Typical Cross Section and General Types of Failures of Impoundment or Reservoir slopes with a Geomembrane Liner System and Cover Soil (after Martin et al. 1984) 8 2.5 Basic Load Transfer Mechanisms (after Jewell et al., 1984) 12 2.6 Bearing Stresses on a Grid Reinforcement During Punching Shear Fail-ure, <f) = 35 (after Jewell et a l , 1984) 14 2.7 Comparison of Test Results with Predicted Values of Bearing Stress (after JeweU, 1990) 15 2.8 Relationship Between Efficiency Factor and Confining Stress for Vari-ous Geogrids (after Juran et al., 1988) 21 3.1 Components of the Pullout Apparatus 30 3.2 Slot Arrangement on the Front Face of the Apparatus 31 3.3 Hopper and the Pullout Test Apparatus 32 3.4 Components of the Hydraulic Control System 34 3.5 Clamp Assembly 36 3.6 Reaction Frame and Air-water Interface Chamber 38 3.7 Location of Pressure Transducers on the Front Face 40 Vlll 3.8 Location of Strain Gauges 41 4.1 Grain Size Distribution Curves for the Sand 44 4.2 Characteristic Dimensions of the Tensar UX1500 Geogrid 46 4.3 Arrangement of the In-isolation Tension Tests 48 4.4 Comparison of Local Strain and Grid Strain at Different Rate of Dis-placement 49 6.1 Pullout Force - Displacement Curves for the Geogrid 58 6.2 Strain Development with Pullout Displacement in Tests G10R05 and G30R05 60 6.3 Strain Development with Pullout Displacement in Tests G50R05 and G90R05 61 6.4 Passive Pressure on the Front face with Pullout Displacement in Tests G10R05 and G30R05 63 6.5 Passive Pressure on the Front Face with Pullout Displacement in Tests G50R05 and G90R05 64 6.6 Volume Change Response for the Geogrid 65 6.7 Comparison of Local Strain Development; SG-1 and SG-5 67 6.8 Comparison of Passive Pressure on the Front Face: TPT-3 & TPT-4 . 68 6.9 Comparison of Passive Pressure on the Front Face: TPT-1 and TPT-6 69 6.10 Pullout Force - Displacement curves for the Geomembrane 71 6.11 Strain Development with Displacement in Tests M05R05 and M08R05 73 6.12 Strain Development with Displacement in Tests M10R05 and M30R05 74 6.13 Passive Pressure Development with Displacement in Tests M05R05 and M08R05 76 6.14 Passive Pressure Development with Displacement in Tests M10R05 and M30R05 77 ix 6.15 Volume Change Response for the Geomembrane 78 7.1 Normalized Pullout Resistance of the Geogrid 81 7.2 Normalized Stress Distribution on the Front face in Geogrid tests . . 84 7.3 Normalized Pullout Resistance for the Geomembrane 85 7.4 Normalized Stress Distribution on the Front Face in Geomembrane tests 87 x List of Symbols B - thickness of the bearing member C - a constant cu - coefficient of uniformity 5^0 - mean particle size of the sand F" - pullout resistance factor Fi - nondimensional stress ratio Fq - embedment bearing capacity factor fb - generalized bond coefficient K - ratio of the actual normal stress to the effective vertical stress Le - embedment or adherence length in the resisting zone - length of the geosynthetic n - number of bearing members Pb - bearing component of pullout resistance Po - maximum pullout load for an isolated bearing member Pr - total pullout resistance Ps - skin friction component of pullout resistance S - spacing between bearing members wT - width of the geosynthetic a - scale effect correction factor Oib - fraction of the bearing area available for bearing af - structural geometric factor for frictional resistance - fraction of the geosynthetic plan area that is solid a/3 - structural geometric factor for bearing resistance xi friction angle between soil and geosynthetic surface average interface friction angle peak interface friction angle local strain in geosynthetic test specimen apparent friction coefficient effective bearing stress acting on the embedded anchor effective normal stress on the geosynthetic surface effective vertical stress at the soil-geosynthetic interfaces average shear stress mobilized along the geosynthetic peak shear stress mobilized along the geosynthetic apparent soil-geosynthetic friction angle xu Acknowledgements I am deeply indebted to Dr. Jonathan Fannin for his unfailing support, encour-agement and patience throughout the study. Without his advice and understanding this project would not have been possible. I would like to express my sincere gratitude to Mr. Art Brookes, technician in the civil engineering workshop, for his help in the development and commissioning of the pullout test apparatus. I would also like to thank Mr. Harold Schemp, Mr. Bernie Merkli, and Mr. Guy Kirsch for their assistance with some parts of the apparatus, Mr. Max Nazar for his assistance with developing the technique of strain gauging the geosynthetic test specimens and Mr. John Wong for his help in setting up data acquistion system. Discussion and co-operation with graduate students Mr. Ralph Kuerbis, Mr. Huaren Dou, Mr. Joyis Thomas, and Mr. M . Uthayakumar regard to the various aspects of experimental work is gratefully acknowledged. This research is funded by an Operating Grant from the Natural Sciences and Engineering Research Council of Canada. I wish to express my thanks for the financial support provided. Thanks are also due to Nilex Geotechnical Products Inc., for supplying Materials used in the testing program. Finally I wish to thank my parents for their continuing love and support. xin To my parents, sisters, and well wishers xiv Chapter 1 Introduction 1.1 Designing with Geosynthetics Polymeric materials such as geotextiles, geogrids and geomembranes that are used in geotechnical engineering applications are collectively termed "geosynthetics". They are used in construction to satisfy one or more of the following basic functions: sepa-ration, reinforcement, filtration, drainage or low permeability barrier. The last decade has seen a tremendous growth in the use of geosynthetics in engineering practice, and given recognition to these materials as an alternative to traditional design solutions. Various factors, such as cost savings, ease of construction and quality control have made the use of geosynthetics attractive. Specifically though, the increasing use has been supported by advances in analytical methods, and may be attributed to • research to evaluate design methodologies and fundamental behaviour; • Standard Tests that facilitate material specifications; • regulatory guidance for construction practice. Design using geosynthetics requires that consideration be given to appropriate analytical methods, material properties, material tests, interpretation of the manu-facturer's technical literature, and soil-geosynthetic interaction. The latter consider-ation, soil geosynthetic interaction is the subject of this research study. 1 Chapter 1. Introduction 2 1.2 Research Needs In the design of reinforced soil structures and membrane lined containment facil-ities, an internal stability analysis examines: • tensile failure of the geosynthetic; • tensile failure of any connections; • pullout failure of the geosynthetic. The objective of this research is to: • design, construct and commission a large pullout apparatus; • take small strain measurements on the geosynthetic test specimens; • better understand soil/geosynthetic interaction as it relates to the mobilization of pullout resistance. Several parameters influence the mobilization of this soil-geosynthetic interaction. The parameters relate to soil type, type of geosynthetic, configuration of the test apparatus, and the testing procedure. While test methods for many material prop-erties, such as tensile strength, have been standardized, there is as yet no standard test method for the measurement of interface shear in a direct shear or a pullout test. This study addresses the general use and interpretation of the pullout test. There is also need for examination of soil-geosynthetic interaction at small strain. Though some significant research contributions have been made, there is an urgent need for quality test data on interface friction between geosynthetics and soil at small relative displacements. This study examines such small strain behaviour through measurements made with strain gauges attached to the geosynthetic test specimens. Chapter 1. Introduction 3 1.3 Thesis Organization The state of the art for pullout testing, and interpretation of the test results is reviewed in Chapter 2. Design and fabrication of the pullout test apparatus used in this research study is described in Chapter 3. Properties of the materials used in the program of testing are reported in Chapter 4. In Chapter 5 the experimental proce-dure followed in the pullout test preparation and performance is described, pullout test results are reported in Chapter 6. In Chapter 7 an analysis and discussion of the results are presented. Some conclusions on the use and interpretation of the pullout test are drawn in Chapter 8. Chapter 2 Literature Review 2.1 Introduction An evaluation of soil-geosynthetic interface friction is important to the design of many structures where the polymeric material is used in an engineering application. In this chapter, the existing literature on applications in soil reinforcement and for low permeability barriers is reviewed, with particular consideration given to factors influencing pullout resistance. 2.2 Technique of Soil Reinforcement While the use of natural fibres to reinforce soils is many thousands of years old, the use of steel and now plastics to strengthen earth structures is somewhat recent. Initially structures were built with galvanized steel strips and a good quality granu-lar backfill, though today geogrids and geotextiles are routinely used as reinforcing elements as well. A distinction that is made in analysis and design is that steel strips are considered inextensible, while polymeric materials are considered extensible. Ex-amples of soil reinforcement applications are shown in Figure 2.1. Placing the reinforcement in a region of tensile strains, and orienting it in the direction of principal tensile strains, will best restrain the tensile stresses and increase the shear strength characteristics of the soil, McGown et al.(1978). The direction of 4 Chapter 2. Literature Review unpaved roads retaining walls reinforcement Figure 2.1: Typical Examples of Soil Reinforcement Applications (after Palmeira, 1987) principal strain is dependent on geometry, construction technique and type of load acting on the structure. The action of reinforcement is mobilized by the stress field in operation, tending to create a composite behaviour in which the "active zone" and the "restrained zone" in the soil mass are bonded, Schlosser (1978). The type of failure surface in a structure will determine which mechanism of soil-reinforcement interaction is developed, Milligan et al. (1987), see Figure 2.2. In the case of failure along surface 1-2, sliding of soil on the plane of reinforcement occurs at A , and the test best suited to model this behaviour is the direct shear test. If failure occurs along surface 3-4, then soil and reinforcement are sheared and the direct shear test, with proper orientation of reinforcement, best models the behaviour. In the case of failure along the surface 5-6, due to insufficient anchorage resistance, sliding of the Chapter 2. Literature Review 6 Figure 2.2: Failure Mechanisms in a Reinforced Soil Retaining Wall (after Palmeira, 1987) reinforcement inside the soil matrix takes place, and the test best suited to model this behaviour is the pullout test. 2.3 Technique of Lining with Geosynthetics Geomembranes are used in the liner systems of waste containment facilities due to their low permeability and chemical compatibility with many waste materials. In some countries, and for certain types of waste, they are used because of regulatory requirements. An example of a liner system is shown in Figure 2.3. Typically the geomembrane liner is taken up the side slope and anchored at the top in a trench. A common failure mechanism of geomembrane-lined side slopes is by slipping of com-ponents within the liner system, Martin et al. (1984), or of the cover soil itself, Seed et al. (1988). A schematic illustration of the mechanism is given in Fig. 2.4. An analysis of slope stability requires: (i) data on limiting shear strength along the interfaces between soil and geosyn-thetic and between different geosynthetic layers; Chapter 2. Literature Review 7 f lame Diamaqe N»tt (7 Lavattl. Covriwt with Smflte lave* ol Balding t ic, 13mm C n n w i i l ' Thtchn#M (ln»talied on Slnpei B#l**»en Pnma«v and Secondary HOPE L m * r i | Nominal 3 * S'oi>» Sand frama^ Blanket f_ J 5 m m (60 M.I | Pnmary L i n " HOPE — Piim«>y L*<lr Of teciion Pwmo 10cm Dia. Petlorated. Wtanped PE 1.5mm (60 M.I) - Secondary Linai-HDPE — Diamagr Venl f ah'ic S Y/MSA C a y Liner \ Secondary Lea* Deieenon Piping • , 0 c n * ° " P t ' , 0 * , , , d - Wfao&ed PE Gtavel Bedding I adding fabric Sue G'ading Fill (Wharr Required) Figure 2.3: Section Through a Liner System (after Udwari and Kittridge, 1986) (ii) an understanding of tension in the liner system on the overall slope stability; (iii) an understanding of slippage between soil, geomembrane and geotextile, and its relationship to the general stress-strain behaviour of the materials. Martin et al. (1984) conducted modified direct shear test on various geosynthetic-geosynthetic and geosynthetic-sand interfaces and concluded that the interface fric-tion mobilized was between 65% and 90% of the peak friction angle for medium dense samples. Eigenbrod and Locker (1987) report the results of direct shear tests results on various geosynthetic and sand samples, performed on both dense and loose sam-ples. Dilation of the dense samples was indicated by mobilization of a distinct peak and residual shear stress. Interface friction mobilized was between 55% and 85% of peak friction angle. Negussey et al. (1989) observed that interface sliding between a geomembrane and granular soils exhibits a peak and residual value, whereas sliding between a geomembrane and geotextile interface does not result in any such peak. Rinne (1989), using the same ring shear apparatus, tested different types of geomem-brane using an angular quartz sand and a rounded sand at high confining stresses. Geomembranes tested were P V C , smooth H D P E and rough H D P E . The soft P V C and harder rough H D P E mobilized a value of interface friction equal to the shear resistance of the sand, for both the angular and the rounded sand. In contrast, the Chapter 2. Literature Review 8 (o) T Y P E S OF SLOPE S T A B I L I T Y F A I L U R E S (b) T Y P E S OF L I N E R S L I P P A G E F A I L U R E S Figure 2.4: Typical Cross Section and General Types of Failures of Impoundment or Reservoir slopes with a Geomembrane Liner System and Cover Soil (after Martin et al. 1984) Chapter 2. Literature Review 9 smooth H D P E mobilized approximately 65% of peak resistance of the rounded sand, and 90% of that for the angular sand. Takasumi et al. (1991) review state-of-the art testing procedures for soil-geosynthetic interface strength characteristics, and con-clude that more testing is required to understand the influence of type and size of the apparatus on interface strength characteristics. The type of failure in a geomembrane-lined structure will determine the mecha-nism of soil-geosynthetic interaction that is developed. In the case of failure at the cover soil surface, or a geosynthetic-geosynthetic surface, the test best suited to model the interaction is the direct shear test. If failure occurs at the anchor trench, then the test best suited to model the interaction is the pullout test. 2.4 Factors Influencing Pullout Resistance Factors influencing pullout resistance are the type of soil, the material properties and geometry of the inclusion, and the configuration of the test apparatus. Soil pa-rameters of interest are the shear strength characteristics, dilatancy, relative density, and the fines content. Since pullout resistance is a function of soil-geosynthetic in-teraction, the parameters of interest are the geometry, the structural aspects of the geosynthetic material (such as in plane or out of plane transverse elements), tensile strength, extensibility, and creep characteristics of the inclusion. The influence of the test apparatus is a result of the loading system, the sample dimensions and its preparation, the boundary conditions and the testing procedure. 2.4.1 Soil Characteristics In construction practice a well-graded free draining granular material is specified for permanent reinforced soil structures. The specification is due to the fact that Chapter 2. Literature Review 10 highly factional materials will develop a better bond with reinforcement than poor materials. A high fines content will tend to restrict the free draining behaviour of a soil, with a consequent potential for decrease in effective stresses. Hence an upper limit to the percentage of fines allowable in the backfill material is usually recom-mended, Brown et al. (1979). This is not to suggest that other soils cannot be used successfully in construction: Murray et al. (1979) used a silty clayey sand as backfill material for a reinforced soil wall and concluded that, despite construction difficulties and pore pressure development, cost-savings could be achieved over granular backfills imported over substantial distances. 2 . 4 .2 Geosynthetic Characteristics Characteristics of a geosynthetic which influence pullout resistance are: tensile strength and extensibility, the effect of confining stress on stiffness, creep behaviour and durability. Geosynthetics are thermovisco-elastic materials, where the load-strain characteristics are dependent not only on strain magnitude but also on strain rate and temperature. McGown et al. (1978) have clearly shown the effect of extensibil-ity by considering extensible and inextensible inclusions. In addition to increasing strength, the principal action of extensible inclusions is to increase soil ductility, and significantly limit any strain softening observed in dense sands. However the action of inextensible inclusions is to increase the strength and deformation modulus but it also to cause the deformation modulus to be more brittle. In comparison with metals, polymeric materials are more deformable and have a large range of tensile modulus. They are particularly susceptible to creep, which is the capacity for continued strain at constant load. As the required service life of a structure may exceed 75 to 100 years for some permanent structures, durability of the geosynthetic and, to some Chapter 2. Literature Review 11 extent the facings, is an important factor. Polymeric materials, although not suscep-tible to corrosion, may degrade due to physico-chemical activity in the soil, such as hydrolysis, oxidation, and environmental stress cracking. Also, they are susceptible to degradation on exposure to ultraviolet light. Creep and durability are accounted for in the design by extrapolating tensile strength data to the design time, and by applying partial safety factors to allow for mechanical and environmental damage, Jewell and Greenwood (1988). 2.4.3 In teract ion between Soi l a n d G e o s y n t h e t i c in P u l l o u t Soil-geosynthetic interaction is an important parameter in the design of a rein-forced soil structure or a low permeability barrier. Three criteria must be satisfied in analysis and design: (i) pullout capacity:- the pullout resistance of the geosynthetic should exceed the design tensile force by a specified factor of safety; (ii) allowable displacement:- the relative displacement between soil and geosynthetic required to mobilize interface shear should be less than the allowable displacement; (iii) long term displacement:- the pullout load should be less than the critical creep load for the geosynthetic. Jewell et al. (1984) refer to the mechanism of interaction for a grid structure in soil as bond, and use a bond coefficient. In a similar approach, Martin et al. (1984) and Eigenbrod and Locker (1987) use an efficiency factor when describing interaction of geomembranes and geotextiles. Soil-reinforcement interaction in direct shear involves some or all of the following general mechanisms of load transfer, see Figure 2.5: (i) lateral friction, where soil shear on plane surface areas of the geosynthetic oc-curs; (ii) passive earth pressure on transverse elements of geogrids, welded wire meshes, Chapter 2. Literature Review 12 a. Shear between soil and plane reinforcement surfaces b. Soil bearing on grid rein-forcement bearing surfaces c. Soil shearing over soil through the reinforcement grid apertures Figure 2.5: Basic Load Transfer Mechanisms (after Jewell et al., 1984) bar mats, and woven geotextiles, developed due to the soil bearing against surfaces normal to the direction of the force to be resisted; and (iii) soil shearing over soil in the apertures of a grid. The transfer of load between soil and geosynthetic in pullout is by mobilization of the first two components only, since there is no relative displacement between soil particles on either side of the element. The two components are lateral friction and passive resistance or bearing. Jewell et al. (1984) derived an expression to describe pullout interaction between grid reinforcement and soil, so that a bond capacity could be calculated from the Chapter 2. Literature Review 13 fundamental properties of the reinforcement geometry and angle of friction of the soil. The skin friction component of pullout resistance for a geosynthetic of length Lr and width WT, placed horizontally, with an effective normal pressure of crn is given by: Ps = 2asLrWra'ntan6 (2.1) where: as is the fraction of the geosynthetic plan area that is solid, LTk,WT are length and width of the geosynthetic, a'n is the effective normal stress on the geosynthetic surface, and S is the friction angle between soil and geosynthetic surface Equation 2.1 is valid for geomembranes and planar geosynthetics without any asper-ities in the third dimension. Passive soil resistance developed against bearing surfaces normal to the direction of force to be resisted is similar to the pressure on the base of deep foundations in soil. Jewell et al. (1984) modified an expression for deeply embedded anchors to establish a theoretical contribution from bearing stresses. A lower bound to the expression is associated with a punching shear failure mode in the soil, (see Figure 2.6), is: ^ = Fi = e(«>+*)"»* tan(45 + ^) (2.2) °n 2 where: <r'b is the bearing stress acting on the embedded anchor; Fi is the non-dimensional stress factor. Figure 2.6: Bearing Stresses on a Grid Reinforcement During Punching Shear Failure, <j> = 35 (after Jewell et al., 1984) Chapter 2. Literature Review 15 1000 100 10 nblo-„ = e*'n* tan' ((jr/4) + (ol2J) 20 0b/oy = e1"2*** * lan ((^/4) + (0/2)) 40 <p: degrees 60 Jewell et al. (1984) O Ovesen & Slroman (1972) •M- Ovesen & Slroman (1972) • Huekel & Kwasniewski (1961) <] Chang el al. (1977) O Jewell (1980) Palmeira & Milligan (1989) 0 • A Present work V Akinmusuru (1978) Audibert & Nyman (1977) Dickin & Leung (1983) Dyer (1985) Peterson (1980) Trautmann & O'Rourke (1985) Wang & Wu (1980) + o Figure 2.7: Comparison of Test Results with Predicted Values of Bearing Stress (after Jewell, 1990) An upper bound value is estimated by taking the conventional characteristic stress. field for a footing rotated to the horizontal, and a horizontal boundary stress in the soil, where: = e*tan*. tan2(45 + - ) (2.3) It was suggested by Jewell et al. (1984) that stress ratio be measured directly in pullout tests, or estimated from curves summarizing test results in the literature. A comparison of the theoretical expressions with experimental data, Jewell (1990), shows good agreement despite the large spread and variability of the test results, see Figure 2.7. The stress ratio is used in the following expression to determine bond Chapter 2. Literature Review 16 coefficient, where: PT = 2LrWr<rn'fitaji<p and tan S 1 -tan <f> + Ctb B 1 S 2 tan 4> (2.4) (2.5) fb = OL where: PT is the total pullout resistance, fb is the generalized bond coefficient, ct^  is fraction of the bearing area available for bearing, B is the thickness of the bearing member, and S is the spacing between bearing member. A limiting value of /{, = 1, that is a fully rough condition, is associated with a limiting value of S. It was pointed out by Milligan et al. (1989) that the derivation of such an expres-sion for bond considered the soil to be a continuum, which is not realistic for coarse grained soils, and while the stress ratio increases significantly when is less than about 15, a safe design is possible when the ratio exceeds 15. Jewell (1990) re-plotted the same results normalizing the stress ratio with respect to the stress ratio where particle size is unimportant. The stress ratio where particle size is unimportant is defined as , when if > 10 and if < 10 "DO <\ oo v'b '20 - B/d50 1 .°~n. oo 10 (2.6) (2.7) This will allow a maximum doubling of the bearing stress due to particle size. Chapter 2. Literature Review 17 The effect of interference between bars of a grid on the bond capacity in pullout tests has been analyzed for geogrid reinforcement by Palmeira (1987). A comparison of the value of the pullout load for a given grid with the value obtained for an ideal grid, denned as one having the bearing pressure of a single isolated member under similar conditions without interference, led to a parameter for degree of interference being defined as DI = 1- A nPQ (2.8) where: Pk is the maximum pullout load for a grid with n bearing members, and P0 is the maximum pullout load for an isolated bearing member of the same grid. Thus an expression to calculate bond coefficient is proposed, where: fb = a, tan 6 tan <f) B IS 1 - DI 2 tan § (2.9) Jewell (1990) suggests further modification to the above relation to account for the first bearing member which acts undisturbed on the sand, and subsequent bearing members for which interference can occur. For a grid with n bearing members, this gives: DI = r i i i _ _ 1 n. S/abB {S/abB (2.10) Where the ratio [~g] is defined as the grid geometry required to achieve a fully rough bond. The FHWA manual of U.S Department of Transportation (Christopher, et al., 1990) for the design of reinforced soil structures recommends the following expression: PT = F*a<rvLeC (2.11) where: Chapter 2. Literature Review 18 Le is the embedment or adherence length in the resisting zone behind the failure surface C is the geosynthetic effective unit perimeter, which is 2 for geosynthetic sheets. F" is the pullout resistance (or friction-bearing-interaction) factor a is a scale effect correction factor o~'v is the effective vertical stress at the soil-geosynthetic interfaces The manual recommends pullout tests be used to determine the pullout resistance factor F". This pullout resistance factor is similar to the bond coefficient proposed by Jewell, et al., (1984). It is given by: F" = Fqap+ Kii'ctj (2.12) where: Fq is the embedment (or surcharge) bearing capacity factor i i ' is a ratio of the actual normal stress to the effective vertical stress; it is influenced by the geometry of the geosynthetic ctp is a structural geometric factor for bearing resistance, where ap = ^ ay is a structural geometric factor for frictional resistance p~ is an apparent friction coefficient for the specific geosynthetic A value of 20 has been suggested for the embedment bearing capacity factor Fq based on available experimental data. For geogrids, geomembranes and geotextiles K = 1 and p.* = tan£ . Due to the extensibility of the geosynthetic, the application of pullout force results in a decreasing shear displacement over the length of the Chapter 2. Literature Review 19 inclusion. Interface shear stress is therefore not uniformly mobilized along the total length of the geosynthetic. A scale effect correction factor a is introduced to consider the extensibility of the material, which is defined as: Tp tan Opeofc where: rav and rp are the average and ultimate interface lateral shear stresses respectively, mobilized along the geosynthetic. 6av and Speak a r e the average and peak interface friction angle respectively mobilized along the geosynthetic. The value of the scale effect correction factor is influenced by strain softening of the compacted granular backfill, extensibility of the material and the length of the geosynthetic. The manual recommends a value of 0.6 in the absence of test results; a realistic value may be in the range 0.6 to 1, with 1 being appropriate for an inextensible material. The factor can be determined by pullout tests performed with different lengths of geosynthetic or derived using analytical or numerical load transfer models which have been "calibrated" through numerical test simulations. 2.5 Laboratory Studies Two commonly used tests for measuring interaction between soil and geosynthetic are the direct shear test and the pullout test. For materials that depend only on skin friction for bond capacity, such as geomembranes and geotextiles, modified direct shear tests are typically used to measure interface shear resistance. Test conditions have been varied to examine the influence of orientation of the test specimen (McGown et a l , 1978; Jewell and Wroth, 1987). However the direct shear test does not apply to Chapter 2. Literature Review 20 situations where the action of bearing is an important component of bond, and does not simulate in any manner the development of strain in pullout loading. While most of the laboratory studies have used granular soils, a few studies have used cohesive materials (Ingold, 1983; Costalonga, 1988). 2.5.1 Pullout Tests The following factors control the mobilization of pullout resistance in the pullout testing of geosynthetics: relative density and dilatancy characteristics of the soil, confining stress, geosynthetic orientation, geosynthetic extensibility, and boundary effects. A synthesis of available experimental data obtained by several investigators has been presented by Juran et al. (1988). Data for grid materials is reported in terms of an efficiency factor (tan tan (f), where (f> is defined as apparent soil-geosynthetic friction angle) for different applied normal stress, see Fig. 2.8. Factors influencing the observed behaviour are discussed below. 2.5.1.1 Relative Density and Dilatancy Characteristics of the Soil Dense soils dilate during shearing, but with confinement from the surrounding soil they experience a restrained dilatancy. The dilatancy characteristics of dense sand and its effect on pullout resistance have been well demonstrated for metallic reinforcements (Schlosser and Elias, 1978; Guilloux et al., 1979). This restrained dilatancy causes an increment of normal stress to act on the element, increasing its pullout resistance. The effect of such dilatancy is dependent on the available confining stress, surface texture of the reinforcement and the density of the soil. Johnston (1985) evaluated the effect of dilatancy by placing pressure cells within the soil sample of a large pullout apparatus to monitor the applied normal stress on Chapter 2. Literature Review 21 6.0 5.5 5.0 ~4.5 i^. 4.0 t 3.5 S t-CJ 3.0 > 2.5 s t j 2.0 U. u. . _ uj 1.5 1.0 0.5 0 Symbol Re fe rence Re in fo rcement S o l i Type Johns tor. (1985) Tensar SR-2 Dense sand • I n f o l d (1983) Tensar SR-1 Dense sand $ - 3 4 ° - 3 5 ° o I n f o l d (1983) N a t i o n 1168 Dense sand 1 - 3 4 ° - 3 5 ° • K o e r n a r (1986) Tansar SX-2 Dense sand / - 4 4 ° Rove et a l . (1985) Tensar SR-2 Loose , F ine sand J*- 3 1 ° - 3 2 ° 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 NORMAL STRESS (PSI) Figure 2.8: Relationship Between Efficiency Factor and Confining Stress for Various Geogrids (after Juran et al., 1988) Chapter 2. Literature Review 22 a Tensar SR-2 geogrid. A normal pressure on dense samples (at peak pullout load) some 1.5 to 3 times higher than the applied vertical stress was measured. It was a result of the top boundary of the apparatus being rigid and restrained against upward displacement. In contrast pullout tests performed on loose samples, see Figure 2.8, show the efficiency factor to be independent of vertical stress. This observation is due to the fact that loose samples tend to contract during testing, resulting in a lower pullout resistance, because there is no restrained dilatancy effect. 2.5.1.2 Confining Pressure The effect of any increase in confining stress is to increase the mobilized pullout resistance, while causing the length of the geosynthetic test specimen that is loaded to decrease (Palmeira, 1987). The importance of confining stress on volume change behaviour of a soil during shear is well recognized. Dense samples at high confining stress behave in a similar manner to loose samples at low confining stress. This phe-nomenon explains the decrease in efficiency factor that is observed for dense samples with increasing vertical stress, see Figure 2.8. 2.5.1.3 Grid Orientation Rowe et al. (1985) and Ingold (1983) report a higher pullout resistance when the transverse element of a grid is placed in the direction of pullout. As more transverse elements are oriented perpendicular to the direction of pullout, and since pullout resistance is mobilized predominantly by passive pressure, the results are as one would expect because of the larger bearing area that is available. Chapter 2. Literature Review 23 2.5.1.4 Geosynthetic Extensibility McGown et al. (1978) have shown that strains vary along the length of extensible inclusions. The effect of extensibility is to mobilize a non-uniform shear stress and shear displacement along the length of an inclusion, with greater shear stress and displacement at the loaded end of a test specimen. Hence, the concept of uniformly mobilized limiting shear resistance is inapplicable for extensible materials. There is need for an appropriate load transfer model which takes into account such extensi-bility. F H W A (1990) recommends the determination of scale effect correction factor, which accounts for extensibility of a geosynthetic from results of pullout tests per-formed on test specimens of different length. 2.5.1.5 Boundary Effects Typically a soil sample for pullout testing is prepared in a rectangular test box with a rigid base and side walls. The top boundary may be rigid or flexible. The influence of a rigid boundary that was free to displace, and a flexible boundary, was examined by Palmeira (1987). A flexible top boundary, typically a surcharge bag filled with water, not only eliminates boundary shear stresses but may be used to measure volume change during the test. For otherwise similar test conditions, use of a flexible top boundary decreases the maximum peak pullout load. Tests performed to examine the influence of roughness of the front wall of a test box showed a marked effect, with a dramatic increase in pullout resistance due to an increase in vertical stress on the sample caused by the frictional component of force on the front wall during pullout. Various techniques have been tried to reduce or eliminate this action. In one approach, Palmeira (1987), polythene sheets with grease in between were used to create a low friction boundary (Palmeira, 1987; Lo, 1990). The influence of such an arrangement is to reduce boundary shear stresses, and Chapter 2. Literature Review 24 hence the complimentary shear stresses on the specimen close to the front face. The measured pullout resistance will be less. In another approach the front boundary was transferred into the soil mass by fixing a sleeve across the width of the pullout box, which embeds in the soil, Bonczkiewicz et al. (1986). To further eliminate the effect of confinement of the geosynthetic and soil within the sleeve, the test specimen was clamped within the sleeve, Juran et al. (1991). Although the introduction of a sleeve solves the problem of front face friction, it complicates the stress distribution within the soil mass at the edge of the sleeve. Juran et al. (1988) report the inconsistency of data on rigid and flexible front faces. More testing is required to clarify the influence of boundary effects. A realistic front boundary condition is intermediate between a rigid and a flexible boundary. There is a need for development of a stress controlled boundary to simulate the actual field condition. Gunther et al. (1991) conducted large scale pullout tests on two types of granular backfills and a clay shale. The clay shale yielded a coefficient of sliding of approximately 60% of the corresponding values for the granular materials. A summary of pullout testing reported in the literature is presented in Table 2.1, with reference to test apparatus, materials and boundary conditions. Several pullout facilities have been designed and used to evaluate the pullout performance of both metallic and geosynthetic reinforcements. The testing facilities are very different in many respects, and make it difficult to compare the pullout performance of different inclusions. 2.6 Field Trials The field trials that are reported in the literature describe the performance of re-inforced soil structures. Chang et al. (1977) report field pullout tests that were performed on steel strips (3mm thick and 60mm wide) placed at different elevations Chapter 2. Literature Review 25 Table 2.1: Summary of pullout Tests (modified from Juran et al., 1988) Location of Box Dimensions(cm) L x W x H Boundary Preparation Type of Materials Sample Preparation Louisiana Slate University (Juran. 1991) 152.41 x60 Top:-Flexib)e air bag, 12in sleeve to minimize passive resistance of the rigid front wall to soil movement during pullout Uniform blasting sand with average grain size 0.2mm and min.and max. densities 1.58 and 1.78 Mg/m3 Compaction coupling vibrations with hammering effect* to simulate field conditions. " University of Alberta (Costalonga, 198ft) 106x36x20 Top boundary-Rigid well graded silty day, Tensar 3R-2& signode TNX-5001, paraGrid dynamically compacted with 4.5kg hammer of height of drop 17.5cm(50% of standard proctor density, 1.49 to 1.58 Mg/m3 Oxford University (Palmeira, 1987) 100x100x100 Top>Flexible bag filled with water Sides and front:-Double layer of polythene with grease in between Leighton Bu2zard Sand 14/25, uniform coarse with a relative density 87%, metallic grids, Tensar SR-1 and SR-2, feoton, Stabilenka 400 air pluviation using a specially designed hopper Asian Institute of Technology, Bangkok (Bergado, 1987)r 100x80x90 Top:-Rigid boundary with pressure applied by means of a hydraulic jack Sandy soil(SC) and clayey soil (CH, Tensar SS2 grids and Bamboo grids. Compacted to 95% standard proctor density at optimum moisture content in 15cm thick layers-Drexel University (Koerner, 1986) -Well graded concrete sand dense samples, 30 mil HDPE, CZ-600 geote-xtile and Tensar SR-2 Vibratory compaction STS (Christopher, 1986) 134.6x66-6x45.7 A sleeve was used to transfer shear load behind the front face to minimize soil arching against the front. FontaineDleau sand uni-formly graded (SP) dense samples, Tensar 332. Signode 250 Prepared by vibratory plate compactor at 91% relative density University of Missouri-Roll* (Lentz, 1986) 76.2x29.2x10.1 Fine sand and coarse iand,Tensar 551 and SR2 Placement of sand using V-shaped hopper in 12.7cm thick layers, to obtain dense samples each layers compacted with vibratory hammer Caltran* (Johnston, 1985) 137x91x51 To reduce friction, the compaction pattern was altered and the sides of the box were watered down prior to loading Medium to coarse sand with some fine sand and iilt(SW), dense samples Tensar SR-2 Compacted to 95% relative density at 1*2% below optimum moisture content. Place in 8-10 inch lifts. Utah State University (Anderson, 1984) 152x183x122 Interaction between the front face and grid was minimized by keeping the transverse wire at least 10cm from the box front Silty sand. Washed sand gravel, dense samples, welded wire mesh Compacted by tamping device near optimum water content (90% in silty and washed sand and 100% in pea gravel), placed in 30cm layers Cambridge University (Jewell, 1980) Prior to sample placement int-erior walls cleaned with inhi-bisol. The walls were lubrica-ted with silicon and covered by a thin membrane. The upper and lower water bags provided a smooth surface. Leighton Bu2zard sand (uniformly graded), dense samples, close coil tension spring, Artificially stiffened tension spring and wel-ded wire steel grid. The sand was prepared by pouring the sand from a specially designed hopper, sand dropped through a flexible tube with a constant drop height of 40cm. University of Grenolobe (Collios, 1981) -Sand, Rounded gravel, Crushed gravel. Ballast stones, and clayd> = 0 Texunionfwoven split film), Bidium U14.U44 and U64, Nortene (geog-rid*) Swedish Geotechnical Institute (Holtz, 1977) 190x70x70 To reduce the side effects, the test specimen was 5cm narrower than the width of the box Tullings sand from Stockholm, G-12 sand from Denmark, medium dense samples, Teknisk vav no. 600( woven poly-ester geotextile) Chapter 2. Literature Review 26 in reinforced earth walls during construction. The fill material was a decomposed granite (sandy gravel, gravelly sand, silty clay and gravel). Two strips were instru-mented with strain gauges on both top and bottom surfaces at 1.5m intervals. The strips were 7m and 14m long, self weight loading induced tensile force for almost the entire length of the strip; however, the externally applied pull load only stressed the 14m long strip up to 3.1m into the fill. There were no additional forces measured beyond this point other than the existing forces induced by the self weight loading be-fore testing. The following conclusions were drawn from the typical load displacement curves: 1. The soil will not be strained significantly until the proportional limit (or yield strength of the inclusion) is reached. At this point, the load-displacement curve becomes nonlinear for the composite steel strip and soil material; 2. The maximum tensile stress in the reinforcement is developed near the front face of the wall for any externally applied force; 3. The required minimum length of steel strip in anchorage is about lm. Bonczkiewicz et al. (1991) performed laboratory and field pullout tests on Mirafi 5T, a continuous filament polyester yarn formed into a biaxial grid by a knitting process, to evaluate stress transfer in geogrids with low junction efficiency. Resistance strain gauges were mounted on geogrid sections to obtain strain data. A pullout rate of 1mm/min was used in testing. Grid displacement was measured using a dial gauge near the face of the wall, and eight strain gauges were used to measure local strains. The behaviour observed during pullout tests in the field was similar to that observed in the laboratory. Chapter 2. Literature Review 27 2.7 Conclusions on Research Needs Pullout resistance and a proper understanding of anchorage, are particularly im-portant to the design of reinforced soil structures and geomembrane lined containment facilities. The following conclusions are drawn: • Our understanding is limited by a relatively small data base. There is a need for a larger data base in order to standardize the pullout testing apparatus. • Our understanding is limited by variations in test equipment and testing proce-dures. Pullout resistance is sensitive to the front face boundary condition, more testing is required to evaluate the influence of a displacement controlled (rigid) and a stress controlled (flexible) boundary on the measured soil-reinforcement interaction. The influence of boundary friction must also be addressed. • There is a need for more experimental data, with associated small strain mea-surement on the geosynthetic test specimen. • There is a need for a unified approach to pullout testing to complement the unified approach to interpretation of pullout test results proposed by the U.S. FHWA. Chapter 3 Apparatus 3.1 Introduction A large pullout apparatus has been designed and constructed at the University of British Columbia for this study. The apparatus is used to evaluate the development of pullout resistance with increasing displacement of a geosynthetic test specimen. The apparatus is described in section 3.2. Instrumentation is used to measure pullout force, pullout displacement, pressure on the front face, and strain in the geosynthetic specimen. The arrangement of instrumentation, and the data acquisition system, is described in section 3.3. 3.2 Large Pullout Apparatus The apparatus comprises several components: a pullout box which contains the soil sample and geosynthetic test specimen; a hopper for controlled placement of soil in the box; an hydraulic system for application of pullout load; a clamp assembly for gripping the geosynthetic test specimen; and a reaction frame. In this section, the design criteria for each of the components are presented. '28 Chapter 3. Apparatus 29 3.2.1 Pullout box The length, width and depth of the pullout box were selected to meet three criteria: 1) the box should be long enough to accept representative geosynthetic test specimens; 2) the box should accept test specimens of length/width ratio up to 2; 3) the width and depth of the box should be large enough to minimize the influence of boundary shear stresses. Geosynthetic test specimens are 1.15m long and 0.63m wide. The internal dimensions of pullout box are 1.3m long, 0.64m wide and 0.63m deep. The box accommodates a sand sample 1.3m long, 0.64m wide and 0.6m deep, with space for a surcharge bag to locate between the top of the sample and the top plate of the box. The pullout box comprises a base frame, base plate, two side frames, and two end plates with supporting frames , see Figure 3.1. The base frame is made of 7.6cm x 7.6cm mild steel tube. It supports the test box and provides reaction for the pullout force and normal stress on the test specimen. The base plate is 1.3cm thick aluminum. It provides a rigid lower boundary for the sample, and supports the side and end frames. The side frames are made of 7.6cm x 7.6cm mild steel tube. A 2.5 cm thick plexiglas sheet, fixed by counter sunk screws to the frame forms the side walls of the test box. An aluminum bar of similar thickness is fixed on top of the plexiglas to prevent damage from the top plate resting on the side wall. A glass sheet, 0.3cm thick, is fixed to the inside face of the plexiglas to reduce side friction in the test box. The front face is made of two aluminum plates, 1.3cm thick, fixed to a frame to give a 12mm slot at the mid-height of the sample. Details of the slot arrangement are illustrated in Figure 3.2. The back face is made of 1.3cm thick aluminum plate. It has a small hole, centrally located, that allows wires from the strain gauge to be taken out of the box. The pullout frame, which supports the pullout assembly used to load the test © S o i l @ Geosynthetic Specimen (3) Base Frame @ Support table for clamp © Clamp (6) Flexible bag filled with water Q) 25.4mm thick MS plate © R e a c t i o n Frame © A n g l e Frame @ LVDT (R) Hydraulic Jack @ Pull-out frame @ LVDT support (ft) Load cell (T§) Pressure control valve © See Detail in figure^ 3.5 O @ Figure 3.1: Components of the Pullout Apparatus Chapter 3. Apparatus 31 ± 12mm Slot T Front Face of the Pull-out box m thick neoprene stripping 12.7mm Figure 3.2: Slot Arrangement on the Front Face of the Apparatus specimens, bolts to the base frame. It is made of mild steel I sections. Bending restraint is provided by stiffeners fixed to the base frame, see Figure 3.1. 3.2.2 H o p p e r A hopper is used for controlled preparation of the sand sample in the pullout box, see Figure 3.3. It comprises an aluminum frame which supports two perforated mild steel plates that are overlapped to create a regular pattern of apertures of constant size. The opening size can be altered to suit the grain size, and therefore type of sand, used in testing. Pneumatically operated cylinders control a trap door that allows the sand to fall from the hopper into the test box below. To prevent dispersion of dust, the sides of the hopper assembly are made of thin plexiglas sheet. The hopper is fixed on legs above the pullout box to give a height of fall in the range 1.4m to 0.8m for the sand sample. Chapter 3. Apparatus 32 Chapter 3. Apparatus 33 3.2.3 Pullout Load Assembly The pullout load assembly comprises an double-acting hydraulic cylinder and as-sociated hydraulic control system, and a clamp which connects the piston of the cylinder to the geosynthetic test specimen. The hydraulic cylinder is bolted to the longitudinal cross-piece of the pullout frame, such that the center of the piston is in alignment with the slot in the pullout box. The hydraulic cylinder, manufactured by Westcoast Cylinder Co., has a 82mm diameter piston with a stroke of 152mm. A control system is used to move the piston of the hydraulic cylinder at a constant rate of displacement. The control consists of an oil pump with a 2HP motor capable of generating 3000psi pressure, and several pressure and flow control valves, see-Figure 3.4. Oil flow to the cylinder is controlled both on inlet and outlet of the cylinder. A variable dump valve mounted on the inlet of the cylinder provides control of delivery pressure. This valve was used during the initial stages of loading to eliminate any instantaneous surge of the piston on starting the pump. Rate of displacement of the piston is controlled by the quantity of oil flowing from the outlet of the cylinder. A miniature needle valve is used to vary the quantity of oil flow to maintain a designated rate of displacement. The actual rate of displacement is determined by the data acquisition system and displayed on screen output. In the initial stages of a test, the rate is maintained by operating the pressure control valve, for the flow control valve set at a predetermined position. Once the maximum pressure on the inlet side of the piston is developed, pullout rate is maintained by operating only the flow control valve. A ball valve, place in parallel with the needle valve, is used to control rapid advance and retraction of the piston. Chapter 3. Apparatus 34 Pressure relief valve Pressure gauge Pressure control valve Pressure gauge 1 - Inlet end of the hydraulic cylinder 2 - Outlet end of the hydraulic cylinder Note: Arrows show the direction of oil flow during testing Figure 3.4: Components of the Hydraulic Control System Chapter 3. Apparatus 35 3.2.3.1 C l a m p The clamp is made of aluminum and comprises three pieces: a lower jawT; an upper jaw; and a central insert. The lower jaw is connected to the piston rod by a self-aligning 9.5cm swivel joint. The inside surface of the lower jaw, which grips geosynthetic during testing, is serrated to provide a good grip. The upper jaw fixes to the lower jaw with four screws. The test specimen is held in the clamp by the central insert. The insert is a wedge shaped bar that bears against a stainless steel rod mounted in the inside face of the upper jaw. The lower face of the insert is serrated. Figure 3.5 illustrates the components of clamp assembly. When the piston is retracted the upper and the lower jaw move as a rigid piece: any attempt by the jaws to open is prevented by G-clamps placed at four locations along the clamp. A geomembrane is held in the clamp by a gripping action alone: any tendency of the insert to move in the jaws causes it to bite into the geomembrane. A geogrid specimen, which is characterized by a regular series of apertures, is held by a different action. The central insert is drilled and tapped to accept a series of studs which seat into the apertures of the grid. Consequently the action of the geogrid in the clamp is one of bearing against these studs. The clamp is supported on low friction delrin rails mounted on a table fixed to the base frame. 3.2.4 Surcharge L o a d i n g Surcharge loading is applied to the sand sample by pressurizing a bag filled with water, see Figure 3.6. The bag 130cm x 64cm x 2.54cm, is made of P V C . The bag is pressurized using an air-water interface chamber that allows measurement of volume change of the sample during the testing. The chamber is a modified triaxial cell. Two ports on the top face serve as a vent to atmosphere while filling the bag, and a pressure inlet to the reservoir during testing. Two ports on the bottom face connect Chapter 3. Apparatus Figure 3.5: Clamp Assembly Chapter 3. Apparatus 37 to the surcharge bag and a water pressure transducer. A regulator on the pressure inlet from the laboratory air supply is used to maintain a constant pressure during the testing. A scale on the transparent body of the reservoir is used to measure water level changes, and determine volume change of the sample. 3.2 .4 .1 Reac t ion Reaction to the surcharge pressure applied to the sand sample is provided by a 2.5cm thick top plate to the pullout box, cross-beams and four tie bars attached to the base frame of the apparatus. The top plate, 1.47m x 0.78m, seats on the pullout box and is bolted in position. The cross beams are 7.6cm x 7.6cm channel sections placed back-to-back, and welded at the ends using 1.3cm thick mild steel plates. The tie bars are 2.5cm diameter high yield strength deformed bars. A schematic illustration of the arrangement is given in Figure 3.6. 3 . 3 Instrumentation Instrumentation is used to measure pullout force, pullout displacement, total pres-sure on the front face of the box, water pressure in the surcharge bag and strain in the geosynthetic test specimen. 3.3.1 P u l l o u t Force Pullout force is measured using a load cell connected between the clamp and hydraulic cylinder. The load cell is manufactured by Interface Inc. and has a 44.5kN(10,0001bs) capacity. It is powered with a 10V DC supply. Chapter 3. Apparatus 38 Figure 3.6: Reaction Frame and Air-water Interface Chamber Chapter 3. Apparatus 39 3.3.2 P u l l o u t Displacement pullout displacement is measured using two linear variable differential transformers (LVDTs) mounted on each end of the central insert of the clamp. They are both DC-DC type SE 373/100, manufactured by SE L A B S , with a 100mm stroke. They are mounted independent of the base frame. Displacement of the insert is taken as the mean of these two measurements, and is used to calculate the rate of displacement of the test specimen. 3.3.3 Pressure on the Front Face Six total pressure transducers are used to measure horizontal pressure on the front face of the test box during pullout testing. The transducers, type A B / H P manufactured by Data Instruments, are bonded semiconductor strain gage pressure transducers. Three transducers have a range 0-100kPa and three have a range 0-50kPa. They are flush mounted on the inside of the front face above and below the slot, see Figure 3.7, to determine the variation of pressure along the depth of the sample. This arrangement does not measure any likely variation of pressure across the front face. 3.3.4 W a t e r Pressure Transducer A gauge pressure transducer, type PT78SP-75 manufactured by Dynisco Co., is used to monitor water pressure in the surcharge bag. The body of the transducer was dismantled, and then assembled under water, to ensure that the cavity is free of air. Full saturation of connecting tubing was ensured by using a hypodermic syringe to fill the tube. The transducer is mounted on a bracket that is fixed at the same level as the surcharge bag. Chapter 3. Apparatus 40 Flexible bag T P T - 1 T P T - 2 T P T - 3 T P T - 4 T P T - 5 T P T - 6 ' f Base Plate All dimensions in mm 632 Figure 3.7: Location of Pressure Transducers on the Front Face 3.3.5 S t r a i n Gauges Strain gauges were fixed on the geosynthetic specimen to measure tensile strain during testing. On geogrids, the gauges were fixed on the longitudinal members, termed ribs, see Fig. 4.2. A full description of the gauges, type EP-08-250BF-350 Option E Manufactured by the Micro-Measurements Division of Measurements Group Inc., and the procedure used for bonding them is given in Appendix-A. Wires from the gauges connect to the data acquisition system through a circuit completion box, where dummy gauges are used to complete a full Wheatstone bridge. Strain measured directly from the gauges is termed local strain. Figure 3.8 illustrates the location of strain gauges on the geosynthetic test specimen relative to the front face of the pullout box. Chapter 3. Apparatus 41 ' SG-I S G - 2 ' • S G - 3 • S G - 4 S G - 5 . ' '. • • — * * 8 0 165mm 1 165 165 | 165 1 4 1 0 i i 1 ' 1150 mm Figure 3.8: Location of Strain Gauges 3.3.6 D a t a A c q u i s i t i o n S y s t e m The data acquisition system consists of a DAS-16 board, a desktop microcom-puter, a signal conditioning unit, and a data acquisition program. The DAS-16 board is a multifunction, high speed A/D(analog/digital) I/O expansion board, manufac-tured by Metrabyte Corp. It includes a 12-bit successive approximation converter, user-selectable gain and accommodates 16 single-ended channels or 8 double ended (differential) channels. The signal conditioning unit was designed and built at U B C . It supplies D.C. input to the transducers, and amplifies the output using a variable gain on each channel. The transducer signals from the conditioning unit are taken to the DAS-16 board which converts the signal from analog to digital. A data acquisition program was written to support the program of testing. Fea-tures of the program are: Chapter 3. Apparatus 42 • Scan before testing:- Continuous scanning of all channels before testing, and recording of initial readings of the transducers. • Scan during testing:- Continuous scanning of all channels during testing, and writing output to hard disk at every lOsec. • Scan during testing:- Two values of the rate of displacement, a lOsec average value and an average rate of displacement since start of the experiment, is written to screen every lOsec. • Scan period:- The default scan time interval of lOsec can be changed at any time during the experiment, and restored at any stage of the experiment Chapter 4 Material Properties 4.1 Introduction Materials used in the program of pullout testing are the soil, geogrid reinforcement and geomembrane liner. Properties of these materials are reported below, based on laboratory testing and manufacturer's technical data. 4.2 Properties of the Sand The soil used in the pullout tests was a dry silica sand, supplied from a source in Washington Valley, Washington, USA, by 0. C. L. Industrial Materials Ltd., Vancou-ver. Grain size distribution curves are reported in Figure 4.1. It is a very uniformly graded sand with little or no fines. The particles are observed to be subangular to angular. Particle size diameters are in the range 0.32mm to 1.18mm, with a value of d50 between 0.7 and 0.9mm. Maximum and minimum void ratio determinations were made on the sand according to A S T M D 2049 - 69. A maximum and minimum void ratio of 1.05 and 0.79 respectively were obtained before commencing the program of testing. Sieve analysis gave a coefficient of uniformity Cu=1.5. Some crushing of the sand grains occurred during pullout testing at pressures of 90kPa. There is evidence of the fraction of smaller particles increasing throughout the test program. After completion of the program of testing, a determination of maximum and minimum 43 Chapter 4. Material Properties 44 Figure 4.1: Grain Size Distribution Curves for the Sand Chapter 4. Material Properties 45 void ratios gave 0.96 and 0.71 respectively. Sieve analysis gave a coefficient of unifor-mity C u=2.1. The implication of this change in gradation of the sand relates to the relative density of the prepared test samples: a characteristic value for the relative density of the test samples is 76%. Any crushing will also tend to the angularity of the sand: while this could lead to a reduction of the impact is likely to be negligible at the low confining stresses used in the program of testing. Two monotonically loaded undrained triaxial tests with pore pressure measurement, indicated a constant volume friction angle of 30.5° for the sand. 4.3 Properties of the Geosynthetics Two types of geosynthetic were used in the program of testing to investigate the influence of structural geometry and tensile strength of the inclusion. The geosyn-thetics used were a uniaxial geogrid reinforcement and a geomembrane liner. 4.3.1 G e o g r i d Reinforcement The uniaxial geogrid was a Tensar UX-1500, manufactured by the Tensar Corp. of Atlanta, Georgia. It is produced from a polyethylene resin, using a process in which punched sheets are stretched to give a specified geometry and strength. Some physical and mechanical properties of the geogrid are reported in Table 4.1 from manufacturer's technical literature. Characteristic dimensions of the geogrid are illustrated in Figure 4.2. Like all polymeric materials, Tensar geogrids exhibit a thermovisco-elastic be-haviour, which is to say the load-strain behaviour at a constant rate of strain is sig-nificantly influenced by the test temperature and strain rate. The long-term design load is reported from interpretation of isochronous load-strain curves, see McGown Chapter 4. Material Properties Table 4.1: Properties of the Tensar UX1500 Geogrid 46 Property Value Interlock Apertures: machine Direction(MD) cross Machine Direction open area Thickness: ribs junctions 14.478cm (nom) 1.676cm (nom) 60% (nom) 0.127cm (nom) 0.432cm (nom) Reinforcement long term design load in MD flexural rigidity tensile modulus in MD Junctions: strength efficiency 31kN/m (min) 4,700,000 mg-cm (min) 1445kN/m (min) 70.45kN/m (min) 90% (min) Material high density polyethylene(HDPE) carbon black 97.5% (min) 2.0% (min) Dimensions roll length roll width roll weight 29.87m 1.0m & 1.31m 0.223kN & 0.290kN Roll length ( Longitudinal)^—«• Figure 4.2: Characteristic Dimensions of the Tensar UX1500 Geogrid Chapter 4. Material Properties 47 et al. (1984). Strain in the geogrid test specimen is reported from gauges mounted on the ribs of the specimen. The measurements are termed local strain. The magnitude of local strain on the rib is different from grid strain, where grid strain is defined as that occurring between the bars of the material. In order to relate local strain to grid strain, some in-isolation tension tests were performed using an Instron tension testing machine. The test arrangement is illustrated in Figure 4.3. Local strain was measured directly with strain gauges; grid strain was deduced from an LVDT mounted between the two adjacent bars of the grid. Tests were performed on a sample 11 ribs wide and 4 bars long. Displacement rates of 0.05, 0.10, 0.25 and 0.5mm/min were used. The relationship observed between local strain and grid strain is illustrated in Figure 4.5. Local strain is smaller than the corresponding grid strain, typically by a factor of two. This aspect of behaviour is generally attributed to the plastic flow of the materials at the node. A dependency of strain magnitude on the rate of displacement during loading is evident in the curves, which is characteristic of a visco-elastic behaviour. 4.3.2 Geomembrane L i n e r The geomembrane used was a Novex smooth H D P E sheet, 60mil thick, supplied by Nilex Geotechnical Products Inc., manufactured using a polyethylene resin. Some physical and mechanical properties of the geomembrane are reported in Table 4.2 from manufacturer's technical literature. Chapter 4. Material Properties Figure 4.3: Arrangement of the In-isolation Tension Tests Chapter. 4. Material Properties 49 2.50 Grid Strain (%) Figure 4.4: Comparison of Local Strain and Grid Strain at Different Rate of Dis-placement Table 4.2: Properties of the Novex Smooth Surface H D P E Geomembrane Property Value Density (min) 0.940 g/cc Minimum Tensile Properties Tensile Strength break(min)) 27MPa Tensile Strength yield (min) 14MPa Elongation Break (min) 700% Elongation Yield (min) 10% Modulus of Elasticity(max) 770MPa Tear Strength 0.2kN Chapter 5 Experimental Procedure 5.1 Introduction and Test Program The pullout resistance of geogrid and geomembrane test specimens has been inves-tigated using a large-scale laboratory pullout test. In total nine tests were performed. Five tests were performed on geogrids, four tests at different surcharge pressures and one for repeatability; four tests were performed on the geomembranes at different sur-charge pressures. Variables monitored during a test were load, displacement, tensile strain, lateral pressure on the front face of the box, and surcharge pressure on the sample. 5.2 Test Preparation Preparation of each test involved embedding the geosynthetic specimen in a con-fined sand sample, and applying a pullout load at a constant rate of displacement. Procedures followed in the preparation of the test apparatus; geosynthetic specimen and sand sample are described in the following sections. 5.2.1 L a b o r a t o r y Test B o x Prior to placement of the sand and geosynthetic, the test box was thoroughly cleaned. The total pressure transducers were then connected to the data acquisition 50 Chapter 5. Experimental Procedure 51 system and a series of initial readings taken with the box empty. The hopper was then lifted and seated on the box. A vacuum cleaner hose was fixed in one plexiglas face of the hopper to remove any fine dust remaining in the air after pouring sand, and the front face slot of the test box was sealed. The laboratory apparatus was then ready for sample preparation. 5.2.2 Geosynthe t ic Specimen Each geosynthetic specimen is cut to the full width of the test box. The total specimen length is 1.32m which, for an embedded length of 1.15m, leaves a protruding length of 0.17m for clamping. Prior to placement in the box, the specimen is strain-gauged. The same routine is used for fixing gauges to a geogrid and geomembrane. Five gauges are fixed, typically along the center line of the specimen, though in some tests one gauge was deliberately offset from the center-line. The step-by-step procedure used to fix a strain gauge is reported in Appendix-A. Three protected wires from each gauge are taken out of the box through a 12mm hole in the back face of the test box. The gauges connect to a wheatstone bridge circuit, which interfaces with the data acquisition system. 5.2.3 Sand Placement Sand was placed in the box at a controlled density by pluviation from the hopper. Placement was made in eight layers, to a finished thickness between 7 and 8cm for each layer. A pour was initiated by release of pressure to two pneumatic cylinders supporting trap doors on the hopper, causing them to retract, and the doors to open. After pouring, the vacuum cleaner was turned on briefly to remove any dust remaining in the box. It was observed that a uniform thickness of loose sand in the hopper did not give a uniform layer in the box, however experience gained in slightly profiling Chapter 5. Experimental Procedure 52 the loose sand soon allowed for a nearly horizontal surface after pluviation. Four pluviated layers brought the sand sample to the mid-height of the slot on the front face of the box. At this point the surface of the sand was levelled to receive the geosynthetic test specimen. Care was taken not to disturb the sand in any significant, wTay. Excess material at any location was removed by scooping, and any low pockets filled by pouring of additional sand by hand through the hopper. The geosynthetic specimen was placed on the surface of the sand with the gauges facing upwards, and the protection tubes for the strain gauge wires were passed through the hole in the back face of the box. Four more layers of sand were then placed following the same procedure. A de-termination of density was made from the mass of sand collected in tins placed at six locations on the sample before pluviation of penultimate layer. The local density thus measured gives the spatial variation for each test. Following placement of the final layer of sand, the hopper was removed and the top surface levelled off as before, in preparation for surcharge loading. 5.2.4 S u r c h a r g e L o a d i n g The sand sample is surcharge loaded using a pressurized P V C bag. It is placed empty, and care is required to avoid pinching it between the top plate and side frames of the box. The top plate and cross-beams are then positioned and fixed to the main reaction frame. A reading of the total pressure transducers on the front face of the box is taken before filling the bag. The bag is then back-filled with water such that the water level in the air-water interface chamber reaches mid-height; during this process the chamber is vented to atmosphere. The water pressure transducer that is mounted in alignment with the surcharging bag is then connected and an initial reading taken. Figure 3.12 Chapter 5. Experimental Procedure 53 shows the surcharge loading assembly and the bag. At this point the venting hose is disconnected from the chamber, and surcharge pressure is applied with control from a pressure regulator on the laboratory air supply. Surcharge pressure is maintained constant during a test by manual adjustment of this regulator. An initial reading on the graduated chamber is taken at the beginning of a test, and any subsequent volume changes noted during the test. 5.3 Pullout Assembly Following application of surcharge loading, the geosynthetic test specimen is at-tached to the pullout assembly. The lower jaw of the clamp is advanced to align with the geosynthetic, the clamp insert and upper jaw are placed on the geosynthetic, and the upper jaw is then bolted to the lower jaw. G-clamps placed on the clamp assembly at four locations prevent any attempt by the jaws to open during pullout loading. 5.4 Test Procedure After applying pressure to the water in the surcharge bag, approximately thirty minutes are allowed for equilibrium to occur: the bag tends to expand to fill the gap between the sample and the top plate, causing the water level in the chamber to fall slightly to a constant value and the water pressure to change accordingly. The pullout test is carried out at a constant rate of pullout displacement of 0.5mm/min. A test is started by setting the main valve supply pressure to the hy-draulic cylinder and simultaneously initiating the data acquisition program. Rate of displacement during pullout testing is controlled by manually operating a pressure control a valve and a flow control valve, see section 3.2.3. At the start of a test, the pressure control valve is opened slowly to allow the supply pressure from the pump Chapter 5. Experimental Procedure 54 to act on the piston without an instantaneous surge. Once the full pressure is de-veloped on the piston, the flow control valve is used to control the rate of pullout displacement. Fifteen channels of data are monitored during a test: six total pressure transducers on the front face; five strain gauges on the geosynthetic specimen; a load cell on the hydraulic cylinder; two displacement transducers on the clamp; and a pressure transducer connected to the surcharge bag. Data is scanned continuously and acquired every lOsecs throughout a test. Pressure in the surcharge bag is adjusted as necessary during testing, and an observation of any volume change made by inspection of the air-water interface chamber. Typically a test is continued to a displacement of 70mm, which for a rate of displacement of 0.5mm/min, gives a time for testing of 2 hours and 20 minutes. 5.5 Post Test Routine At the end of the test the hydraulic cylinder is stopped and the pump switched off, surcharge pressure released, and water allowed to drain out of the bag. Al l the instrumentation cables are disconnected from the data acquisition unit. The reaction frame is the dismantled, the top plate lifted off the box, and the surcharge bag taken out. Sand is removed from the box using a modified vacuum cleaner. A measurement of the mass of the storage drums before and after emptying the test box is used to determine the mass of sand in the box, and hence calculate the mean density of the sample knowing the volume of the box. A typical routine for testing is to clean the box and strain gauge a geosynthetic specimen on day 1. On day 2 the box is filled and instrumentation connections made. Surcharge load is applied and test carried out on day 3. On day 4 the sand is removed from the box. A typical testing routine requires 4 days of work for each test. Chapter 6 Results 6.1 Introduction The pullout resistance of the geosynthetics is described from measurements of pullout force and displacement, strain in the geosynthetic, water pressure in the surcharge bag, and pressure on the front face of the test box. Parameters maintained constant in the test program were specimen size and pullout rate. Each specimen was 1.15m x 0.63m and was loaded at a constant rate of 0.5mm/min. Parameters varied in test program were surcharge pressure, and type of geosynthetic. Tests were performed for surcharge pressures in the range 5 to 90 kPa. Two types of geosynthetic, a grid and a membrane, were tested. In total nine tests were performed, five on grid specimens and four on membrane specimens. A summary of test program is reported in Tables 6.1 and 6.2. The tabulated vertical stress is the vertical effective stress on the test specimen. A reference code is established to identify each individual test, where: G 05 to R05 R M 90 I type of geosynthetic(grid & membrane) 1 pull-out rate I 0.5mm/mln) repeatability test surcharge pressure (kPa) Chapter 6. Results 56 Table 6.1: Geogrid Test Parameters Test Test Glob. Density Loc. Density Vertical Stress code Sequence Mg/mz Mg/m3 kPa G10R05 2 1.452 1.423 12.6 G30R05 3 1.450 1.431 28.0 G50R05 4 1.462 1.446 54.4 G90R05 1 1.456 1.435 88.8 G50R05R 9 1.474 1.469 55.5 6.2 Results of Tests on the Geogrid In the early tests, control of the rate of pullout displacement was maintained using only a flow control valve on the outlet of the hydraulic cylinder. Difficulties were experienced in achieving a uniform rate of displacement at the very start of a test, as evident in G90R05 and G10R05 tests. Placement of a pressure control valve on the inlet of the cylinder greatly improved the hydraulic control system and allowed a uniform rate of displacement to be maintained in the remaining tests. 6.2.1 Dens i ty The density of a sand sample is controlled by the height of fall during pluviation. Global density is reported from the total mass and volume of sand in the box. Local density is reported as the mean of the six measurements taken from tins placed on the sand surface prior to pouring the penultimate layer. Typically the values of local density are less than the corresponding global density. While some variation may be expected, because the drop height is less at the location of local density measurement, it is believed this variation indicates sensitivity of the density to rate of pouring. The global density of the sand samples is seen to increase during the program of testing. The likely reason is the occurrence of particle crushing during testing at Chapter 6. Results 57 higher confining stress. The resultant particle degradation tended to increase the fine contents of the sand, encouraging a closer packing of particles. A comparison of particle size distribution curves for the sand before and after the program of testing, see Figure 4.1, supports this postulated behaviour. A representative value of relative density for the sand samples is 76%, based on measurements of global density, and considering mean of values maximum and minimum densities determined according to A S T M procedure D 2049-69. 6.2.2 P u l l o u t Force The development of pullout force on the geogrid test specimens with increasing displacement, at different applied surcharge pressures, is shown in Figure 6.1. The test code indicates the nominal effective stress on the test specimen. The actual value of stress is given in Table 6.1. Pullout resistance is characterized by increasing force with displacement. A common early response is observed at small displace-ments. Thereafter pullout resistance is related to the magnitude of confining stress, with higher surcharge pressures giving a greater mobilized pullout force. The curves illustrate the difficulties experienced in controlling the rate of pullout displacement which varied between 0.35mm/min to 0.70mm/min in tests G10R05 and G90R05, particularly G90R05 in which force is seen to vary considerably. Test G10R05 appears to have mobilized a maximum pullout resistance, given a nearly constant value of force with increasing displacement. In contrast the other tests show a general trend toward such a constant value of force with increasing pullout displacement, but have not yet attained it. Test G90R05, performed at a very much higher surcharge pressure than other tests, shows small strain at the embedded end, see Fig. 6.13, suggesting that the test specimen is tending toward failure in tension rather than pullout. Chapter 6. Results 58 Figure 6.1: Pullout Force - Displacement Curves for the Geogrid Chapter 6. Results 59 6.2.3 L o c a l S t r a in Local strain is denned as that recorded directly from the strain gauge fixed to the rib of the geogrid specimen. The development of local strain with increasing displacement is illustrated in Figures 6.2 and 6.3. Typically local strain is characterized by increasing strain with increasing pullout displacement. Variations from this typical response for tests G10R05 and G90R05 again show the difficulties experienced in maintaining a constant rate of displacement. Gauge SG-1 is located nearest to the clamped end of the specimen, with the others in sequence and gauge SG-5 located nearest to the embedded end, see Figure 3.8. The curves show a consistent response, with largest strains observed near the clamped end of the specimen, and smallest strains near the embedded end of the specimen. Strain propogation along the length of the sample is illustrated by the larger strains recorded at SG-1 compared to SG-5 at any particular displacement. In tests G90R05 and G10R05, the decrease in local strain during initial stages of the test is due to poor control of the rate of pullout displacement. Variations of this characteristic response are caused by failure of the gauges. Local strains up to 2% are measured before debonding of the gauge occurs. In this type of failure the gauge gradually debonds and the strain reading returns to the original value, confirming the integrity of the gauge and its connections. The loss of bond occurs between the adhesive and the geogrid surface. Failure of the gauges in test G10R05 was different and caused by the ripping of the tabs. In this test, a single strand wire w7ith a protective tubing of small diameter was used. The wire was relatively inflexible and the tubing did not properly accommodate movement of the wire, resulting in excessive stress on the solder tab. Al l five gauges failed in this test. Chapter 6. Results 2.50 2.00 -g 1.50 c co CO 3 i.oo Q 60 0.50 -0-00 f i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i I i 0 10 20 30 40 50 60 70 Displacement (mm) 2.00 1.50 c a CO n o o 1.00 0.50 -0.00 G30R05 »-«-. SG-2 SG-3 >» > » > SG-4 ooeoo S6-5 | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I 0 10 20 30 40 50 60 70 Displacement (mm) Figure 6.2: Strain Development with Pullout Displacement in Tests G10R05 and G30R05 Chapter 6. Results 61 Displacement (mm) Displacement (mm) Figure 6.3: Strain Development with Pullout Displacement in Tests G50R05 and G90R05 Chapter 6. Results 62 6.2.4 Passive Pressure on the Front Face Passive pressure is reported from total pressure transducers mounted on the front face of the box. The locations of the transducers are shown in Figure 3.7. The development of passive pressure with increasing displacement is illustrated in Figures 6.4 and 6.5. A general trend is evident: the characteristic response one of increasing of passive pressure with the increasing displacement. The rate of increase and the magnitude of pressure depends on the position of the transducer in relation to the geogrid test specimen. Transducers TPT-3 and TPT-4, which are nearer to the specimen, show the greatest pressure increase in all tests except G10R05. Transducers TPT-1 and TPT-6, which are most distant from the geogrid sample, show the least response in terms of magnitude and rate of increase of pressure. An intermediate response is shown by TPT-2 and TPT-5. The values of horizontal stress measured before pullout loading may be used to establish a coefficient of earth pressure. Some variation was observed in the calculated values. A characteristic value for the samples given by the mean for all transducers in all tests is 0.43. This would suggest an angle of shearing resistance of 35°. 6.2.5 V o l u m e Change The volume change measurements give an indication of the vertical deformation of sand. They appear to be significantly dependent on the surcharge pressure applied to the sample, see Figure 6.6. It is evident from the curves that the lower the confining pressure, the greater is the volume of water expelled from the bag. The sand sample is prepared in a relatively dense state. Pullout of the test specimen causes volumetric expansion to occur in the sand adjacent to the sample in those tests at lower confining stresses. While at higher confining stress, some volumetric contraction occurs. Chapter 6. Results 63 100 7 5 -(0 ! , 50-G10R05 < < » » » TPT-1 TPT-2 • - M - a - a TPT-3 > > > > > TPT-4 O O O O O TPT-5 Dttftlttf- TPT-6 0 10 20 30 40 50 Displacement (mm) 60 70 100 G30R05 « . H » - » * » • » - » » « I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I 0 10 20 30 40 50 60 70 Displacement (mm) Figure 6.4: Passive Pressure on the Front face with Pullout Displacement in Tests G10R05 and G30R05 Chapter 6. Results 64 G50R05 I I I I | I I I I | I II I | I I I I | I II I | I I I I | I I I I 0 10 20 30 40 50 60 70 Displacement (mm) i I I i | i i i i | i i i i | i i i i | i i i i | i i i i | i i i i 0 10 20 30 40 50 60 Displacement (mm) 70 Figure 6.5: Passive Pressure on the Front Face with Pullout Displacement in Tests G50R05 and G90R05 Chapter 6. Results 65 2000 - 2 5 0 --500-|—i—i—i—i—|—[—r—i—i—| i—i—i—[—|—r~i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—r~ 0 10 20 30 40 50 60 70 Displacement (mm) Figure 6.6: Volume Change Response for the Geogrid Chapter 6. Results 66 6.2.6 R e p e a t a b i l i t y To check repeatability of the test procedure, one geogrid test was performed for conditions very similar to test G50R05, see Table 6.1. Results from the repeatability test are compared for mobilized pullout resistance, strain developed along the geogrid specimen and development of passive pressure on the front face. The mobilized pullout resistance compares very well at small displacement, see Figure 6.1, and reasonably well at larger displacement. This difference in measured force is in part the result of the lkPa higher vertical effective stress on test specimen G50R05R. The development of local strain for gauge SG-1 and SG-5 is shown in Figure 6.7. There is very good agreement between the SG-1 measurements in the two tests: in both cases gauge SG-1 failed by debonding. There is reasonable agreement between the SG-5 measurements, particularly at large displacements. Gauge SG-5 was still functioning at the end of the test. The development of passive pressure on the front face of the test box at location TPT-3 and TPT-4 is shown in Figure 6.8. The magnitude and the rate of increase of pressure with displacement agree reasonably well for location TPT-4. However measurements at location TPT-3 show considerable difference in both magnitude and rate of increase of pressure. The two transducers are equidistant from the slot in the front face, and should give a similar response during testing. In fact both TPT-4 readings and TPT-3 reading for test G50R05R indicate this similar response. The lower values measured at location TPT-3 for test G50R05 are likely caused by the test specimen being placed slightly below the mid-height of the slot: a comparison of the sum of pressures measured at TPT-3 and TPT-4 locations in each tests shows little difference. Measurements taken at locations TPT-1 and TPT-6, see Figure 6.9, are farther away from the geogrid test specimen and therefore less influenced. They Chapter 6. Results * 67 1.50 Displacement (mm) Figure 6.7: Comparison of Local Strain Development; SG-1 and SG-5 Chapter 6. Results 68 0 | i i i i | i i i i | i i i i | i i i I | I I I I | I I I i | i i i i 0 10 20 30 40 50 60 70 Displacement (mm) Figure 6.8: Comparison of Passive Pressure on the Front Face: TPT-3 &: TPT-4 suggest a good repeatability of measured response. 6.3 Results of Tests on the Geomembrane Four tests were performed on the geomembrane specimens. A summary of the test parameters has been presented in Table 6.2. The magnitude of surcharge pressure applied in testing was limited to a nominal value of 30kPa, to avoid tensile yield of the geomembrane specimens. Chapter 6. Results 69 35 z » » » » » G50R05 G50R05R I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I 10 20 30 40 Displacement (mm) 50 60 70' Figure 6.9: Comparison of Passive Pressure on the Front Face: TPT-1 and TPT-6 Table 6.2: Geomembrane Test Parameters Test Test Glob. Density Loc. Density Vertical Stress code sequence Mg/m3 Mg/m3 kPa M05R05 5 1.467 1.446 6.5 M08R05 7 1.467 1.460 7.6 M10R05 8 1.467 1.453 12.3 M30R05 6 1.468 1.467 30.7 Chapter 6. Results 70 6.3.1 Dens i ty Determinations of global density of the sand sample show that it remained con-stant for all tests on geomembranes. The likely reason for this phenomenon is the low confining stresses used in this series of testing, which did not cause any particle crushing to occur. As reported for tests on the geogrids, values of local density are typically less than the corresponding global density. 6.3.2 P u l l o u t Force The mobilization of pullout force on the geomembranes with increasing displace-ment, at different surcharge pressures, is shown in Figure 6.10. The test code indicates the nominal effective stress on the test specimen. The actual value of stress is given in Table 6.2. A typical response of increasing force with increasing displacement is ob-served. The response is consistent with level of confining stress, with higher confining stress mobilizing a greater pullout resistance. Most curves tend to a constant value of force at large displacement. Test M05R05 is an exception to this pattern of behaviour, showing a slow decrease in pullout force with increasing displacement. This was the first test to be performed on a geomembrane test specimen, and the behaviour is due to slipping of the test specimen in the clamp. The clamp was modified to effect a good grip by cutting serrations on the lower jaw and central insert which are in contact with the geomembrane. No slip of the test specimen in the clamp was observed thereafter. Tests M08R05 and M10R05 illustrate the characteristic response of pullout force increasing with displacement to a constant or nearly constant value which is asso-ciated with full mobilization of pullout resistance. Test M08R05 achieves this full mobilization earlier in the test than M10R05, because the mobilization of a limit interface friction is progressive along the test specimen. In test M30R05, the pullout Chapter 6. Results 71 25 Displacement (mm) Figure 6.10: Pullout Force - Displacement curves for the Geomembrane Chapter 6. Results 72 force is close to the yield strength of the material and the test was stopped when dis-tress of of the test specimen was observed. Yielding of the specimen at the clamped end was causing the clamp to tilt, and there was evidence of impending tensile failure of the specimen. 6.3.3 L o c a l S t ra in The development of tensile strain with increasing displacement is illustrated in Figures 6.11 and 6.12. Typically local strain is characterized by increasing strain with increasing pullout displacement. Strain propogation in the specimen is again apparent from a comparison of strain at location SG-1 with other locations through to SG-5. Variations from this typical response for test M05R05 support the observation of slip of the test specimen in clamp initiating at a displacement of nearly 8mm, given the decrease in strain at all locations of measurement. Considering the response of the other tests, test M08R05 shows increasing strain to a displacement of some 30mm, and a nearly constant strain thereafter. A typical strain gauge failure by ripping of the wires is evident at location SG-2. To check the load distribution across the specimen, strain gauge l a was deliberately offset by 150mm from the centre line parallel to SG-1 in test M10R05. The strain measurement on both the gauges, shown in Figure 6.12, indicates uniform load distribution across the width of the specimen. Both the gauges have failed by debonding but at different strain levels. Test M30R05 shows large strains near the clamped end of the specimen, and very small strains near the embedded end. Pullout load is mobilized over a relatively short length of the specimen near the front face of the test box. The response is in good agreement with the impending tensile failure of the specimen between the clamp and the box observed during testing. Chapter 6. Results 73 c j= 53 15 u o 0.60 0.50 0.40 -0.30 0.20 -0.10 -0.00 j | | | | | | | | | | | | | | | | | I I | | | | | | | | | | | | | i i 0 10 20 30 40 50 60 70 Displacement (mm) 2.5 2.0 g 1-5 H c 'co CO o o 0.5 -0.0 0 10 20 30 40 50 60 70 Displacement (mm) Figure 6.11: Strain Development with Displacement in Tests M05R05 and M08R05 Chapter 6. Results 74 M10R05 » < » » » SG-1 SQ-1a Displacement (mm) Figure 6.12: Strain Development with Displacement in Tests M10R05 and M30R05 Chapter 6. Results 75 6.3.4 Passive Pressure on the Front Face The development of passive pressure on the front face of the box with increasing displacement is presented in Figures 6.13 and 6.14. The response of the transducers nearer to the test specimen, at locations TPT-3 and TPT-4, shows similar response of horizontal pressure above and below the slot, although of different magnitudes. Transducers TPT-1 and TPT-6, at the top and bottom boundary, show very little response to pullout loading. It would appear that the sand specimen is largely undis-turbed during the complete duration of the test, except for a thin zone adjacent to the test specimen. A peak value of pressure on the face is observed at displacements that compare well with the peak pullout force in all the tests. Values of earth pressure coefficient established using the same approach as that reported for the geogrid tests indicated significant variation. The measured response is likely due to the low confining pressure that was applied in the geomembrane tests. 6.3.5 V o l u m e change The relationship between volume change and pullout displacement for tests on the geomembrane is shown in Figure 6.15. Again the curves show that a greater volume of water is expelled from the surcharge bag at lower values of confining stress. A marked difference between the response during tests on the geomembranes and tests on the geogrids is that volume change is very small in all of the tests except for test M05R05. The observed volume change behaviour is as one would expect because the zone of soil influenced in the mobilization of pullout force is smaller. This is confirmed by the measurements of total pressure on the front face of the test box. Chapter 6. Results 30 76 M05R05 * * * * * TPT-1 ' i ' I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I 10 20 30 40 50 60 Displacement (mm) 70 40-M08R05 TPT-1 TPT-2 TPT-3 TPT-4 I I I I | I M I | I I I I | I I I I | I I I I | I I. I I | I I I I 0 10 20 30 40 50 ' 60 70 Displacement (mm) Figure 6.13: Passive Pressure Development with Displacement in Tests M05R05 and M08R05 Chapter 6. Results 77 30-M10R05 oo^eo TPT-5 * * * * * TPT-6 l | l l l l | i I i I | I I I I | I I I I | I 20 30 40 50 60 Displacement (mm) 70 100 • M30R05 * * * * * TPT-1 TPT-2 TPT-3 TPT-4 ooooo TPT-5 * * * * * TPT-6 0 | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I | I I I I 0 10 20 30 40 50 60 70 Displacement (mm) Figure 6.14: Passive Pressure Development with Displacement in Tests M10R05 and M30R05 Chapter 6. Results 78 Figure 6.15: Volume Change Response for the Geomembrane Chapter 7 Analysis of Results 7.1 Introduction Results of the pullout tests on geogrid and geomembrane test specimens are an-alyzed and discussed separately in this chapter. Analysis is made with respect to bond coefficients, or efficiency factors as they are sometimes termed, that are used in design practice. The approach is consistent with that adopted by the U.S. FHWA (1990) document for design of reinforced soil structures. Specifically the development of force and strain in the test specimens, and the associated horizontal stresses on the front face of the box, are examined. Discussion examines the significance of results with respect to behaviour of the polymeric test specimens. The practical implications of the test results are reviewed. 7.2 Pullout Resistance of the Geogrid Pullout resistance of a grid material is mobilized by direct shearing of soil over the surface of the grid, and bearing of the grid members in the soil. The magnitude of pullout force at any displacement is a function of grid geometry, normal effective stress, and the friction angle of the soil. During testing at the constant rate of pullout displacement used in this program, the applied vertical stress on the test specimen 79 Chapter 7. Analysis of Results 80 and width of the specimen are constant. The mobilization of pullout force is normal-ized with respect to normal effective stress on the specimen and embedded surface area, and plotted against displacement in Figure 7.1. With respect to the approach adopted by the U.S. FHWA (1990), and summarized in section 2.4.3, equation 2.11, the nondimensional form is given by Similarly, with respect to the approach proposed by Jewell et al. (1984), and sum-marized in section 2.4.3, equation 2.4, the nondimensional form is given by A plot of the relationship between F'a and displacement, see Figure 7.1, shows the following: • a smooth development of bond factor with the increasing pullout displacement; • the parameter F'a varies between 0 and a maximum value approximately 0.9; • a nearly constant value of F'a in test G10R05 that indicates impending pull-out failure,given the mobilized tensile force is less than the long-term tensile strength; • test G90R05 indicates the specimen is tending toward failure in tension rather than by pullout, because F'a is increasing with pullout displacement at the end of the test; • test G30R05 would appear to be tending toward pullout failure, while tests G50R05 and G50R05R would appear to be tending toward tensile failure; and • very good repeatability for the two tests performed at similar surcharge pres-sures. Pr = fb tan (j) (7.2) Chapter 7. Analysis of Results 81 F*a Displacement (mm) Figure 7.1: Normalized Pullout Resistance of the Geogrid Chapter 7. Analysis of Results 82 The curves are plotted for the embedded length changing with pullout displace-ment during testing., assuming an essentially inextensible behaviour, where change in embedded length is the pullout displacement. The inextensible behaviour assumed for reduction of the test results gives an upper bound value to the parameter F'a, since the assumed embedded length is shorter than actual embedded length. An embedded length that is unchanging with pullout displacement would be compatible with a fully extensible behaviour, and therefore would give a lower bound to the parameter F'a., since the assumed embedded length is longer than the actual embedded length. A dual data point marked for the last point of each curve establishes the sensitivity of this bond factor to the embedded length. The likely behaviour of geogrid is between the range of inextensible and extensible response and is governed by confining stress. At higher confining stress, where the response is inextensible, the lower bound value better represents behaviour of the test specimens. At lower confining stress, where the response is extensible, the upper bound value represents behaviour of the test specimens. The U.S FHWA design guidance recommends a factor to account for the extensibility of polymeric materials, a, which varies in the range 0.6 and 1. A value of 0.6 is recommended in the absence of test results. The extensible behaviour is supported by strain gauge measurements, particularly in those tests where strain at location SG-5 is very small at the end of a test, see Figure 6.3. Local strain measurements at location SG-1 are approximately in the range of 2 to 3%. This is equivalent to a maximum grid strain approximately in the range of 4.5 to 7% given the relationship identified between local strain and grid strain in section 4.3.1. A performance limit strain of 10% is reported by the manufacturer, beyond which creep failure of the material rapidly occurs. Independent measurements of strain and force from an instrumented sloped reinforced soil, Fannin and Hermann (1990), show working grid strains in the range 1 to 2%, for mobilized forces in the Chapter 7. Analysis of Results 83 range 1 to 4 kN/m. The laboratory test results greatly exceed these working limits. In a similar approach to the evaluation of pullout resistance, Jewell et al. (1984) propose a bond coefficient fb, by combining equations 7.1 and 7.2, which is: 7b tan c6 = F " a (7.3) Typical values for this bond coefficient are between 0.8 and 1.0. A maximum value for fb tan <f> of approximately 0.9 would suggest a friction angle for the sand in the range 48° to 42° respectively. This range is not unreasonable for an angular coarse grained sand in a dense state at low confining stresses. Horizontal stress increment on the rigid front face is normalized with respect to the mean shear stress on the geogrid specimen, and reported in Figure 7.2. The mean shear stress is computed by considering pullout force at the end of the test and the embedded area of the test specimen. The data are not necessarily the maximum pressure readings, rather they are those at the end of the test. The magnitude of the stress ratio is highest at those locations close to the geogrid test specimen. It is interesting to note the difference in pressure distribution above and below the test specimen: somewhat larger stresses are observed on the front face below the slot, though in general the bottom face is less stressed than the top face. A possible reason for this behaviour is the loss of sand through the slot. A similar response was reported by Palmeira (1987). 7.3 Pullout Resistance of the Geomembrane Pullout resistance of a. geomembrane is mobilized by direct shearing of soil over the surface of the sheet. The mobilization of pullout force is again normalized with respect to normal effective stress on the specimen and embedded surface area, and plotted against displacement in Figure 7.3. Dual data points are reported at the - 0 . 8 0 -- 1 . 0 0 -Figure 7.2: Normalized Stress Distribution on the Front face in Geogrid tests end of a test in a similar manner to that for the geogrid tests. The curves show the following: • a relatively smooth development of bond with increasing pullout displacement except for M05R05 due to the reason given below; • the parameter F'a varies between 0 and a maximum value approximately 0.9; • clamp slippage in test M05R05; • test M30R05 indicates tensile failure of geomembrane, where the stress near the clamped end was close to yield stress of the material; and • pullout type failure in tests M08R05 and M10R05 indicated by constant value of F'a. Chapter 7. Analysis of Results 1.00 0.80 H 0.60 0.40 H 0.20 H 0.00 i i i i | I I I I | I I I I | I I I I | I I I I 20 30 40 50 60 70 Displacement (mm) Figure 7.3: Normalized Pullout Resistance for the Geomembrane Chapter 7. Analysis of Results 86 The maximum pullout resistance measured in tests M08R05 and M10R05 sug-gests an interface friction angle of 41° and 36° respectively, if extensibility of the geomembrane is not considered and therefore a equals unity. Test M10R05 on the geomembrane may be compared with test G10R05 on the geogrid, since the test con-ditions are similar. The maximum bond factor F*a in test M10R05 is approximately 0.70, while in test G10R05 it is approximately 0.91. It would appear that the grid structure is more efficient in mobilizing a bond with the soil. Strains measured using the strain gauges are those strains in the geomembrane test specimen at the location of the gauge: the concept of local and grid strain is not applicable for such sheet materials. The magnitude of strain measured in the embedded specimens is in the range 0 to 2%, and significantly less than the yield strain of the geomembrane which is reported by the manufacturer as 10%. Extensibility of the membrane is evident in test M30R05 performed at a large magnitude of confining vertical stress, where measurement of strain at SG-5 indicates zero strain throughout the test. The horizontal stress increment on the rigid front face during testing is normalized with respect to the mean mobilized shear stress on the geomembrane test specimen, and reported in Figure 7.4. The data are not necessarily the maximum pressure readings, rather they are those at the end of the test. The magnitude of stress ratio is considerably less than that measured in the geogrid tests, with the exception of TPT-3 reading in test M30R05. The negative value of stress ratio for test M05R05 is caused by the horizontal stress at the end of the test being less than the initial value. It is a result of some loss of sand adjacent to the slot during pullout, and slippage of the test specimen in the clamp which caused the increment of horizontal stress during pullout to be small. The lower ratio near the slot for test M10R05 is likely due to loss of sand through the slot during testing. In test M30R05 where resistance to pullout Chapter 7. Analysis of Results 1.00 -i 87 • • • • • M05R05 A A A A A M08R05 A A A A A M10R05 M30R05 - 0 . 2 0 H - 0 . 4 0 H A ( T T - 1 . 0 0 J Figure 7.4: Normalized Stress Distribution on the Front Face in Geomembrane tests was mobilized in a short length near to the slot, a very high stress ratio is observed. Chapter 8 Conclusions A large pullout test apparatus has been designed and built to load geogrid and geomembrane test specimens in a sand sample. A series of tests have been performed to commission the apparatus and examine the influence of confining stress on pullout resistance. Based on interpretation of the test results, the following conclusions are drawn: 8.1 On Mobilization of Pullout Resistance 1. Mobilized pullout force is significantly influenced by confining stress. For similar test conditions, a grid structure is more efficient in mobilizing bond than a planar sheet of the same type of polymer. 2. The implication of the embedded length for anchorage in reinforced soil walls and anchorage trench design in waste containment facilities is apparent from pullout response. The resistance to pullout loading greatly exceeds typical val-ues. 3. A progressive development of strain is observed in the test specimens. The development is governed by level of confining stress, with greater strain propa-gation taking place at lower confining stresses. 88 Chapter 8. Conclusions 89 4. For the geogrid reinforcement, the allowable long-term design load is typically mobilized at small strains approximately 1%. 5. Our understanding of behaviour in the pullout precludes its development as a Standard Test procedure. 8.2 On Improvements to the Test Procedure 1. Measurement of horizontal stress during testing show a response that is sensitive to the position of the geosynthetic test specimen in the slot. Care should be taken in control of pluviation of the sand sample. 2. Some crushing of the sand grains was observed during testing at high confining stress. A sand of hard grains that is less susceptible to crushing is best-suited to these test conditions. 3. Improvement to the sand placement technique with a better control over rate of pouring. 4. A closed-loop hydraulic feedback system will give improved displacement control during testing and allow for load control during testing. 8.3 On Suggestions for Future Work 1. Further development of the apparatus is required to standardize the equipment, and move toward interpretation of the pullout test as an elemental test. 2. More testing on specimens of different length is required to properly evaluate a, the scale effect correction factor which accounts for extensibility of the material. Chapter 8. Conclusions 90 3. A suitable load transfer model needs to be developed for interpreting pullout test results of extensible inclusions. 4. The influence of the front boundary condition on pullout resistance needs further evaluation: a stress controlled boundary may better reproduce conditions in the field. 5. The mechanisms by which dynamic stress increments are resisted requires ex-amination for design of structures subject to earthquake loading. References [1] Abdel-Motaleb A. A. (1989), "Pull-out Resistance of Welded Wire Mats in Clayey Silt Backfill," Utah State University, Logan, thesis submitted in partial fulfillment of M.S. degree. [2] Bathurst, R. J. , and Jarrett, P. M . (1987), "Large-Scale Models Tests of Geo-composite Mattresses over Peat Subgrades," Transportation Research Record 1188, Washington, D.C., pp. 28-36. [3] Bergado, D. T., Bukkanasuta, A. , and Balasubramaniam, A. S. (1987), "Lab-oratory Pull out Tests Using Bamboo and Polymer Geogrids Including a Case Study," Geotextiles and Geomembranes, Vol. 5, No. 3, pp 153-189. [4] Bonczkiewicz, C , Christopher, B. R., and Atmazidis, D. K. (1986), "Evaluation of Soil-Reinforcement Interaction by Large-Scale Pull-out Tests," Transporta-tion Research Record 1188, Washington, D.C., pp. 1-18. [5] Bonczkiewicz, C , Christopher, B. R., and Simac M . (1991), "Load Distribution in Geogrids With Low Junction Efficiency," Proc. Geosythetics '91, Atlanta, Georgia, February 1991, Vol. 2, pp. 643-652. [6] Brand, S. R., and Duffy, D. M . (1987), " Strength and Pullout Testing of Geogrids," Proc. Geosynthetic '87, New Orleans, Feb., Vol. 2, pp. 226-236. [7] Brown I. E. W., and Rochester, T. A., (1979), "Reinforced Earth - Technical Guidance Provided by the Department of Transport, England," Proc. Int. Conf. on Soil Reinforcement, Paris, Vol. 2, pp. 423-430. [8] Chang, J . C , Hannon, J . B., and Forsyth, R. A. (1977), "Pull resistance and in-teraction of earthwork reinforcement and soil," Transportation Research Record 640, Washington, D.C., pp. 1-7. [9] Christopher, B. R., Gil l , S. A. , Giroud, J . P., Juran, I., Mitchell, J . K. , Schlosser, F. and Dunnichff, J. , (1990), "Design and Construction Guidelines for Rein-forced Soil Structures - Volume I," and "Summary of Research - Volume II," Report No. FHWA-RD-89-043, Federal Highway Administration, U.S Depart-ment of Transportation. [10] Collios, A. , Delmas, P., Gourc, J . P., and Giroud, J . P. (1980)," Experiments on . Soil Reinforcement with Geotextiles, " ASCE National Convention, Portland, Oregon, pp. 53-73. 91 References 92 [11] Costalonga M . A. R. (1988), "Geogrid Pull-out Tests in Clay," The University of Alberta, thesis submitted in partial fulfillment for the degree of M.Sc. [12] Eigenbrod, K. D., and Locker, J. G. (1987), "Determination of Friction Values for the Design of Side Slopes Lined or Protected With Geosynthetics," Canadian Geotechnical Journal, pt, pp. 509-519. [13] Fannin, R. J., and Hermann, S. (1991), "Performance Data for a Sloped rein-forced Wall," Canadian Geotechnical Journal, 27(5), pp. 676-686. [14] Gunther, E. B., Abd E l Halim, A. 0., and Shelton, W. S. (1988), "Pull-out Resistance of Geogrids Embedded in Granular Aggregates," Proc. 3rd Canadian Symposium on Geosynthetics, Kingston, Ontario, pp. 91-98. [15] Gunther, E . B., Halim, A. 0. A. , and Shang, Q. (1991), "Large-Scale Pullout Tests: Assessment of Procedure and Results," Proc. Geosythetics '91, Atlanta, Georgia, February 1991, Vol. 2, pp. 615-627. [16] Holtz, R. D. (1977), "Laboratory Studies of Reinforced Earth using a Woven Polyester Fabric," Proc. Int. Conf. on the use of Fabrics in Geotechniques, Paris, Vol. 1, pp. 149-154. [17] Ingold, T. S. (1983), " Laboratory Investigation of Grid Reinforcement in Clay," Geotechnical Testing Journal, Vol. 6, No. 3, pp. 112-119. [18] Ingold, T. S., and Miller K . S. (1988), "Geotextiles Handbook," Thomas Telford Limited, England. [19] Jewell, R. A. , Milligan, G. W. E., Sarsby, R. W. and Dubois, D. D. (1984), "Interaction Between Soil and Geogrids," Proceedings of the Symposium on Polymer Grid Reinforcement in Civil Engineering, London, England, pp. 18-30. [20] Jewell, R. A. (1985), "Material Properties for the Design of Geotextile Rein-forced Slopes," Geotextiles and Geomembranes, Vol. 2, pp. 83-109. [21] Jewell, R. A., and Wroth, C. P. (1987), "Direct Shear Tests on Reinforced Sand," Geotechnique, 37, pp. 53-68. [22] Jewell, R. A. , and Greenwood, J . H. (1988), "Long Term Strength and Safety in Steep Soil Slopes Reinforced by Polymer Materials," Geotextiles and Geomem-branes, Vol. 7, No. 1, pp. 81-118. [23] Jewell, R. A. (1990), "Reinforcement Bond Capacity," Technical note, Geotech-nique, Vol. 40, No.3, 513-518. References 93 [24] Johnston, R. S. (1985), "Pullout Testing of Tensar Geogrids," University of California at Davis, California, thesis submitted in partial fulfillment of MS degree. [25] Juran, I., and Chen, C. L. (1987), "Soil-geotextile Pullout Interaction Proper-ties: Testing and Interpretation," Transportation Research Record 1188, Wash-ington, D.C., pp. 37-47. [26] Juran, L , Knochenmus, G., Acar, Y . B. , and Arman, A. (1988), "Pull-out Response of Geotextiles and Geogrids (Synthesis of Available Experimental Data)," Proc. Symposium on Geosynthetics for Soil Improvement, Tennessee, May 1988; (edited by R. D. Holtz.) pp. 92-111. [27] Juran, I., Farrag, K. H. , and Richmond, L. (1991), "Short and Long Term Performance of Polymeric Geogrids," Proc. Geosythetics '91, Atlanta, Georgia, February 1991, Vol. 2, pp. 587-599. [28] Lentz, R. W., and Pyatt, J . N . (1987), "Pull-out Resistance of Geogrids in Sand," Transportation Research Record 1188, Washington, D.C., pp. 48-55. [29] Lo, S. C. R , "Discussion on Scale and Other Factors Affecting the results of Pullout tests of Grids Buried in Sand," Geotechnique, 40, No.3, pp. 519-521. [30] Martin, J . P., Koerner, R. M . , and Whitty, J . E. (1984), "Experimental Fric-tion Evaluation of Slippage Between Geomembranes, Geotextiles and Soils," International Conference on Geomembranes, Denver, U.S.A. pp. 191-196. [31] McGown, A., Andrawes, K . Z., and A l Hasani, M . M . (1978), "Effect of Inclu-sion Properties on the Behaviour of Sands," Geotechnique, Vol. 28, No. 3, pp. 327-346. [32] Milligan, G. W. E., and Palmeira, E. P. (1987), "Prediction of Bond between Soil and Reinforcement," Prediction and Performance in Geotechnical Engi-neering, Calgary, June 1987, pp. 147-153. [33] Mitchell, J . K. , and Villet, W. C. B. (1987), "Reinforcement of Earth Slopes and Embankments," National Cooperative Highway Research Program Report No. 290, Transportation Research Board, Washington, DC. [34] Murray, R. T., and Bolden, J . B. (1979), "Reinforced Earth Wall Constructed with Cohesive Fi l l , " Proc. 3rd Int. Conf. on Geotextiles, Vienna, Vol. 3, pp. 701-706. [35] Murray, R. T., Carder, D. R., and Krawczyk, J . V . (1980), "Pullout Tests on Reinforcement Embedded in Uniformly Graded Sand Subject to Vibration," Proc. 7th European Conference on Soil Mechanics and Foundation Engineering, Brighton, Vol. 3, pp. 115-120. References 94 [36] N A T O Advanced Research Workshop on the "Application of Polymeric Rein-forcement in Soil Retaining Structures," (1987: Kingston, Ont. ) The applica-tion of polymeric reinforcement in soil retaining structures - editors, Peter M. Jarrett, Alan McGown. p. cm.-(NATO ASI series. Series E, Applied sciences ; no. 147). [37] Negussey, D., Wijewickreme, W. K. D., and Vaid, Y . P., (1989), "Geomembrane Interface Friction," Canadian Geotechnical Journal, Vol. 26(1), pp. 165-169. [38] Palmeira, E. M . (1987), "The Study of Soil-Reinforcement Interaction by means of Large Scale Laboratory Tests," University of Oxford, thesis submitted in partial fulfillment for the PhD degree. [39] Palmeira, E. M . , and Milligan, G. W. E. (1989), "Scale and Other Factors Affecting the Results of Pull-out Tests of Grids Buried in Sand," Geotechnique, 39, No. 3, pp. 511-522. [40] Ralph, H. , Walls, J. , and Carrol, R. G. (1986), "Geogrid Reinforcement of Granular Bases in Flexible Pavements," Transportation Research Record 1188, Washington, D. C , pp. 19-27. [41] Ramalho-Ortigao, J. A., and Palmeira, E. M . (1982), "Geotextile Performance at an Access Road on Soft Ground Near Rio de Janeiro," Proc. 2nd Int. Conf. on Geotextiles, Las Vegas, USA, Vol. 2, pp 353-358. [42] Rinne, N . (1989), "Evaluation of Interface Friction Between Cohesionless Soil and Common Construction Materials," The University of British Columbia, thesis submitted in partial fulfilment of requirements for M.A.Sc degree. [43] Rowe, R. K. , Ho, S. K. , and Fisher, D. G. (1985), "Determination of Soil-Geotextile Interface Strength Properties," Proc. Canadian Symp. on Geotextiles and Geomembranes, Edmonton, September, pp. 885-890. [44] Rowe, R. K. , and Ho, S. K. (1986), "Determination of Geotextile Stress-Strain Characteristics using a Wide Strip Test," Proc. 3rd Int. Conf. on Geotextiles, Vienna, pp.885-890. [45] Saxena, S. K. , and Budiman, J . S. (1985), "Interface Response of Geotextiles," Proc. of the 11th Int. Conf. on Soil Mechanics and Foundation Engineering, San Franscisco, CA, Vol. 3, pp. 1801-1804. [46] Schlosser, F., (1978), "History, Current and Future Developments of Reinforced Earth," Proc. Symp. on Soil Reinforcing and Stabilising Techniques, N.S.W. Institute of Technology, pp. 5-28. [47] Sluimer, G., and Risseeuw, P. (1982), "A Strain Gauge Technique for Measuring Deformations in Geotextiles," Proc. 2nd Int. Conf on Geotextiles, Las Vegas, U.S.A, Vol. 3, pp. 835-838. References 95 [48] Sweeney, B. P., and Clough, G. W. (1990), "Design of a Large Calibration Chamber," Geotechnical Testing Journal, Vol. 13, No . l , pp. 36-44. [49] Takasumi, D. L., Green, K. R., and Holtz, R. D. (1991), "Soil-Geosynthetic Interfaces Strength Characteristics: A Review of State-of-the-Art Testing Pro-cedures," Proc. Geosythetics '91, Atlanta, Georgia, February 1991, Vol. 1, pp. 87-100. [50] Udwari, J . , and Kittridge, J . (1986), "Designing of Double Lined Impoundments-Lessons Learned," Proc. 3rd Int. Conf. on Geotextiles, Vienna, Austria, Vol. 3, pp. 929-934. [51] U K Department of Transport (1988), "Performance of a Reinforced Earth Bridge Abutment at Sirling," Transport and Road Research Laboratory, Re-search Report No. 128. [52] Vaid, Y . P., and Negussey, D. (1988), "Preparation of Reconstituted Sand Spec-imens," Advanced Triaxial Testing of Soil and Rock, A S T M STP 977, American Society for Testing and Materials, Philadelphia, pp. 405-417. [53] Weiler, W. A. , and Kulhawy, F. H. (1982), "Factors Affecting Stress Cell Mea-surements in Soil," Journal of the Geotechnical Engineering Division, ASCE, Vol. 108, No. GT12. pp. 1529-1548. Appendix A Technique of Strain Gauging Plastics A . 1 I n t r o d u c t i o n The development of a strain gauging technique for plastics requires that consider-ation be given to the mechanical, thermal and chemical properties of theses polymeric materials. With regards to mechanical properties, plastics have a relatively low mod-ulus of elasticity in comparison to metals. Consequently there is potential for large strain magnitudes, which place demands on the capacity for elongation of a strain gauge, the adhesive, and the wiring procedures. In addition, any tendency of the strain gauge to impart an effect of local reinforcement to the test specimen must be recognized. With regard to thermal properties, polymeric materials have thermal coefficients approximately 5-10 times greater than those of metals and concrete. The thermal conductivity of the plastics influences both the selection of gauge size and excitation voltage to achieve an acceptable power dissipation per unit of grid area; it also increases the difficulty of maintaining an active and dummy strain gauge at the same temperature in a variable thermal environment. With regard to chemical properties, care must be taken to avoid any reaction between the geosynthetic test specimen and those chemicals used as cleaning solvents, adhesives, and protective coatings for the gauges. Consideration of these factors is important to the selection of a high-elongation strain gauge for measuring relatively large strains, a suitable surface preparation 96 Appendix A. Technique of Strain Gauging Plastics 97 Table A . l : Dimensions of the Strain Gauge Gauge Length in mm Overall Length in mm Grid Width in mm Overall Width in mm Matrix Size(mm) L x W 6.35 9.53 3.18 3.18 13.2 x 5.6 technique for the test specimen, compatible solvents and cleaning agents, and an adhesive to achieve a good bond given an acceptable curing time. A.2 Characteristics of the Strain Gauge The strain gauge selected for the program of laboratory testing is type EP-08-250BF-350 Option E, manufactured by the Micro-Measurements Division of Mea-surements Group Inc. It is selected for the following reasons: • the EP series gauges are made of a special annealed constantan foil with a tough high elongation polymide backing that offers high elongation capacity; • the geometry of the gauge, defined by the gauge pattern designation 250BF and reported in Table A - l . l , fits well on the ribs of the geogrid test specimens; • a high resistance gauge minimizes heat dissipation, for which the 350 ohm is selected; • encapsulation of the gauge, the option E, protects the gauge circuit from damage by abrasion with the backfill sand. Appendix A. Technique of Strain Gauging Plastics 98 A.3 Strain Gauging Procedure A.4 Chemicals for Surface Preparation A 1,1,1 Trichloro-ethane solvent is used to degrease the surface of the test spec-imen because of its inertness to polyethylene. The degreaser prevents embedment of contaminants in the surface of the geosynthetic specimen. A No. 400 grit paper is used to roughen the surface for bonding. It is an important factor in getting a good bond between the polymide backing of the strain gauge and the polyethylene material. The surface is then neutralized with a mild ammonia solution, which leaves it with a slightly alkaline pH. Gauge installation is performed within few minutes of completing the surface conditioning. A.5 Adhesive Selection Cyanoacrylate adhesive M-bond 200 is selected because of the simple, quick curing procedure. Although the bond is brittle at large strain, for this test series where relatively small strain measurements are important, the measurable strain that occurs before debonding is acceptable. A.5 .1 Geosynthe t ic Surface P repa ra t i on Supplies required: 1,1,1 trichloro-ethylene degreaser, 400 grit sand paper, gauze sponge, compressed air, cotton swab, and M-Prep neutralizer 5. Steps: 1. Trim the geosynthetic test specimen to the required dimensions. Secure it on clean flat surface and mark the gauge locations. Precise alignment of the gauge with the direction of loading is important for meaningful data. Appendix A. Technique of Strain Gauging Plastics 99 2. Spray the gauge location with 1,1,1 trichloro-ethylene degreaser and wipe clean using gauze sponge. 3. Use No. 400 grit sand paper to roughen up the surface, sanding first at a 45° angle to the direction of testing and then at right angles to get a pattern of cross hatches. Approximately 4 minutes of sanding is required. 4. Using compressed air, clean the gauge location to remove any small particles. 5. Neutralize the surface by wiping the location with M-Prep Neutralizer 5, a mild ammonia solution, which leaves it with a alkaline pH. 6. The gauge should be applied within 2 or 3 minutes of completing the surface preparation. A . 5 . 2 Gauge P repa ra t ion Supplies required: Plexiglas frame(rectangular hollow),1,1,1 trichloro-ethylene degreaser,tweezers, eraser, MGJ-2 tape, and strain gauges. Steps: 1. Clean the plexiglas frame with the 1,1,1 trichloro-ethylene degreaser, wiping with a gauze sponge. 2. Take a small length of MJG-2 tape and tape it down on the plexiglas frame causing the tape to be exposed at the hollow portion. 3. Remove the strain gauge from its package, ensuring it is held on the edge using tweezers. 4. Place the gauge on to the exposed tape, aligning it parallel with the edge of the tape. Low air pressure from the compressed air is used to affix it firmly. Appendix A. Technique of Strain Gauging Plastics 100 5. The gauge is now ready for transfer to the geosynthetic test specimen. A . 5 . 3 A p p l i c a t i o n of the G a u g e Supplies required: M Bond-200, catalyst 1-1-1 Trichloro-ethylene, gauze sponge, TFE-1 sheet, silicone pad, aluminum block, and MJG-2 tape. Steps: 1. Lift the tape off the plexiglas frame along with the gauge and attach it to the geosynthetic at the desired gauge location, aligning the gauge in the testing direction. The tape on the side of the terminal should not be pressed firmly, but the opposite side should be. 2. Peel back the tape from the terminal side at an acute angle so that the tape lifts off with the gauge. Pull back the tape 3mm further than the edge of the gauge. 3. Coat the base of the gauge with catalyst 1-1-1 trichloro-ethylene, taking care not to apply excess material. Allow sufficient time for the catalyst to dry, since it will not cause the adhesive to bond in a wet condition. 4. Apply two drops of M Bond-200 adhesive, a cyanoacyrlate ester, to the geosyn-thetic test specimen at the gauge location and quickly lower the gauge to make contact. 5. Using the gauze sponge, apply a uniform pressure to the gauge. 6. Overlay the gauge with TFE-1 film, a silicone pad, and a aluminum block, and apply thumb pressure for 1 minute. 7. Remove thumb pressure and allow 15 minutes to pass before peeling off the tape. Appendix A. Technique of Strain Gauging Plastics 101 8. Carefully peel off the tape from the terminal side, pulling back at an angle of more than 150°. 9. The gauge is ready for soldering. A . 5 . 4 G a u g e So lder ing Supplies required :Rosin solvent, 3-strand wires, and soldering accessories Steps: 1. Cut the 3-strand wire into desired lengths, and pass it through the stiff plastic tubing that is used to protect the wire from damage by the sand grains. 2. Solder the ends of the wires and trim to leave 2mm exposed. 3. Tape down the stiff tubing to the geosynthetic test specimen forming a loop of excess wire adjacent to the gauge. 4. Brush the gauge surface with rosin solvent to remove dust particles. 5. Using flux and solder, and taking care not to apply excess heat that will damage the test specimen, quickly place solder on the tabs of the gauge. 6. Solder the prepared wires to the solder on the gauge tabs. 7. Check the resistance of the gauge and its connection using an ohm-meter. 8. Clean the surface with rosin solvent to remove flux. 9. The gauge assembly is now ready for protecting. Appendix A. Technique of Strain Gauging Plastics 102 A . 5 . 5 G a u g e P r o t e c t i o n Supplies required: Cellophane tape, M-coat A, TFE-1 and MJG-2 tape. Steps: 1. Coat the gauge assembly with M-coat A, a polyurethane coating, placing three coats at an interval of 30 minutes. 2. Coat the exposed wires between the gauge and protective tubing as well. 3. Cover the gauge assembly with TFE-1 film, a Teflon film, and tape it down firmly using cellophane tape or MJG-2 tape. Corrections which are to be applied to the measured data are: transverse sensi-tivity; thermal output; gauge factor variation with temperature; Wheatstone bridge nonlinearity; and gauge factor variation with strains. Considering all these factors, the measured percentage strain in a full bridge circuit is related to the change in electrical output recorded, by the following expression: A.6 Analysis of Strain Data %e = 4 £ 0 (A. l ) (F + e)10 - 2 * E0(F + e) where: E0 is the output of the bridge in mV E is the input to the bridge in V , F is the gauge factor supplied by the manufacturer. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0062786/manifest

Comment

Related Items