Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evaluation of interface friction between cohesionless soil and common construction materials Rinne, Norman F. 1989

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

Item Metadata

Download

Media
831-UBC_1989_A7 R56.pdf [ 4.08MB ]
Metadata
JSON: 831-1.0062826.json
JSON-LD: 831-1.0062826-ld.json
RDF/XML (Pretty): 831-1.0062826-rdf.xml
RDF/JSON: 831-1.0062826-rdf.json
Turtle: 831-1.0062826-turtle.txt
N-Triples: 831-1.0062826-rdf-ntriples.txt
Original Record: 831-1.0062826-source.json
Full Text
831-1.0062826-fulltext.txt
Citation
831-1.0062826.ris

Full Text

EVALUATION OF INTERFACE FRICTION BETWEEN COHESIONLESS SOIL AND C O M M O N CONSTRUCTION MATERIALS by NORMAN F. RINNE B.A.Sc, University of Waterloo, 1985  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES Department of Civil Engineering The University of British Columbia  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA MARCH, 1989 © NORMAN F. RINNE, 1989  In  presenting this  degree at the  thesis  in  University of  partial  fulfilment  of  of  department  this thesis for or  by  his  or  requirements  British Columbia, I agree that the  freely available for reference and study. 1 further copying  the  representatives.  an advanced  Library shall make  it  agree that permission for extensive  scholarly purposes may be her  for  It  is  granted  by the  understood  that  head of copying  my or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department  of  CiiA  t  i q f p»TiA  The University of British Columbia Vancouver, Canada  DE-6 (2/88)  ABSTRACT A reliable quantitative assessment of interface friction parameters between construction materials and the surrounding soil mass where this interface represents a potential failure surface, will allow for less conservatism and/or safer design of soilstructures. This can only be achieved if the factors affecting the interface friction angle 8, are adequately understood.  The purpose of this thesis is to obtain such an  understanding. The influence of soil particle shape, confining pressure and amount of relative displacement on the value of 6 are studied in the laboratory using a ring shear apparatus. Two quartz sands, one angular and the other rounded, with steel, concrete and geosynthetics were used as the interfacing constituents. Test data indicate that the value of 8 can vary significantly for each of the surfaces investigated. Smooth HDPE geomembrane exhibits distinct peak and residual 8 values that range from 65 to 90% of the friction angle of the surrounding sand. Rough HDPE and PVC geomembrane interfaces are shown to mobilize the full friction angle of the sand. Steel surfaces display a complex interface frictional response that is strongly affected by the amount of relative displacement along the interface. However, concrete surfaces mobilize essentially identical 8 values at small and large displacements that are approximately equal to the <J>CV of the interfacing sand.  ii  CONTENTS PAGE ABSTRACT LIST OF FIGURES ACKNOWLEDGEMENTS  ii iv vii  1. INTRODUCTION  1  2. INTERFACE FRICTION - LITERATURE REVIEW  4  2.1 2.2 2.3 2.4 2.5 2.6  General Sand-Steel Interface Testing Sand-Concrete Interface Testing Sand-Geomembrane Interface Testing Surface Texture Assessment Summary  3. EXPERIMENTATION  15  3.1 Ring Shear Testing - General 3.2 UBC Ring Shear Apparatus 3.2.1 Testing Configurations 3.3 Test Materials 3.3.1 Sands 3.3.2 Construction Materials 3.3.2.1 Steel 3.3.2.2 Concrete 3.3.2.3 Geomembranes 3.4 Testing Program  15 15 17 18 18 21 21 22 23 24  4. GEOMEMBRANE INTERFACE TEST RESULTS 4.1 4.2 4.3 4.4  4 6 8 8 10 14  PVC Smooth HDPE Rough HDPE Comparison Of Surface Performance  5. STEEL INTERFACE TEST RESULTS 5.1 Smooth Steel 5.2 Rough Steel 5.3 Comparison Of Surface Performance  27 27 34 51 55 58 58 63 67  6. CONCRETE TEST RESULTS  74  7. CONCLUSIONS  79  REFERENCES  iii  LIST OF FIGURES PAGE Figure 2.1  Typical Modified Direct Shear Apparatus For Interface Testing  5  Figure 2.2  Components of Surface Texture  11  Figure 2.3  Surface Roughness Amplitude Parameters  13  Figure 3.1  UBC Ring Shear Apparatus  16  Figure 3.2  Ottawa and Target Sand Grainsize Distribution Photomicrograph of Ottawa and Target Sand Grains  Figure 3.3  19 20  Figure 4.1  Interface Friction vs. Shear Displacement Ottawa Sand - PVC Interface (0 - 25 mm)  28  Figure 4.2  Interface Friction vs. Shear Displacement Target Sand - PVC Interface (0 - 25 mm)  29  Figure 4.3  Roughness Coefficient vs. Normal Stress After Shearing Ottawa Sand - PVC Interface  31  Figure 4.4  Interface Friction vs. Shear Displacement Ottawa Sand - PVC Interface (0 - 25 mm)  32  Figure 4.5  Interface Friction vs. Normal Stress Ottawa Sand, Target Sand - PVC Interface  33  Figure 4.6  Interface Friction vs. Shear Displacement Ottawa Sand - HDPE Interface (0 - 25 mm)  35  Figure 4.7a  Interface Friction vs. Shear Displacement Target Sand - HDPE Interface (0 - 25 mm)  37  Figure 4.7b  Interface Friction vs. Shear Displacement Target Sand - HDPE Interface (0 - 350 mm)  38  Figure 4.8  Interface Friction vs. Normal Stress Ottawa Sand, Target Sand - HDPE Interface  39  Figure 4.9  Typical Surface Roughness Profiles Ottawa Sand, Target Sand - HDPE Interface  40  iv  Figure 4.10  Photomicrograph of Grooves Ploughed in HDPE  Figure 4.11  Roughness Coefficient vs. Normal Stress Ottawa Sand, Target Sand - HDPE Interface  Figure 4.12 Hard Asperity Ploughing Through Soft Surface Figure 4.13  42  43 45  Peak Interface Friction vs. Normal Stress Ottawa Sand - HDPE Interface Residual Interface Friction vs. Normal Stress Ottawa Sand - HDPE Interface  48  Figure 4.15  Peak Interface Friction vs. Normal Stress Target Sand - HDPE Interface  49  Figure 4.16  Residual Interface Friction vs. Normal Stress Target Sand - HDPE Interface  50  Figure 4.17  Interface Friction vs. Shear Displacement Ottawa Sand - Rough HDPE Interface (0 - 25 mm)  52  Figure 4.18  Interface Friction vs. Shear Displacement Target Sand - Rough HDPE Interface (0 - 25 mm)  53  Figure 4.19  Peak Interface Friction vs. Normal Stress Ottawa Sand, Target Sand - Rough HDPE Interface  54  Figure 4.20  Peak Interface Friction vs. Normal Stress Ottawa Sand - PVC, HDPE Interfaces  56  Figure 4.21  Peak Interface Friction vs. Normal Stress Target Sand - PVC, HDPE Interfaces  57  Figure 5.1a  Interface Friction vs. Shear Displacement Ottawa Sand - Smooth Steel Interface (0 - 25 mm)  59  Figure 5.1b  Interface Friction vs. Shear Displacement Ottawa Sand - Smooth Steel Interface (0 - 650 mm)  60  Figure 5.2a  Interface Friction vs. Shear Displacement Target Sand - Smooth Steel Interface (0 - 25 mm)  61  Figure 5.2b  Interface Friction vs. Shear Displacement Target Sand - Smooth Steel Interface (0 - 650 mm)  62  Figure 5.3  Roughness Coefficient vs. Normal Stress Ottawa Sand, Target Sand - Smooth Steel Interface  64  Figure 5.4a  Interface Friction vs. Shear Displacement Ottawa Sand - Rough Steel Interface (0 - 25 mm)  65  Figure 4.14  v  46  Figure 5.4b  Interface Friction vs. Shear Displacement Ottawa Sand - Rough Steel Interface (0 - 650 mm)  66  Figure 5.5a  Interface Friction vs. Shear Displacement Target Sand - Rough Steel Interface (0 - 25 mm)  68  Figure 5.5b  Interface Friction vs. Shear Displacement Target Sand - Rough Steel Interface (0 - 650 mm)  69  Figure 5.6  Roughness Coefficient vs. Normal Stress Ottawa Sand, Target Sand - Rough Steel Interface  70  Figure 5.7  Peak Interface Friction vs. Initial Surface Roughness Ottawa Sand, Target Sand - Steel Interfaces  71  Figure 5.8  Residual Interface Friction vs. Initial Surface Roughness Ottawa Sand, Target Sand - Steel Interfaces  72  Figure 6.1  Interface Friction vs. Shear Displacement Ottawa Sand - Concrete Interface (0 - 25 mm)  75  Figure 6.2  Interface Friction vs. Shear Displacement Target Sand - Concrete Interface (0 - 25 mm)  76  Figure 6.3  Photomicrograph of Exposed concrete Aggregate  77  vi  ACKNOWLEDGMENTS I would like to express my thanks to Dr. Y.P. Vaid for his invaluable support and guidance throughout this research program. To Art Brooks and Harold Schemp I would like to convey my genuine appreciation for your talents as machinists. Your assistance with the equipment was of great help to the successful completion of this research program. I am very grateful to Dr. P. Ko and Mark Robertson at N.R.C. for providing access to the profiling equipment in the Tribology laboratory. Their generous assistance enabled this research program to draw upon data that otherwise would have been unavailable. Many thanks to National Seal and Gundle Linings for supplying the geomembranes used in this study.  vii  1. INTRODUCTION The mobilized friction angle associated with displacement along the interface of soils and construction materials is a major component in a wide variety of geotechnical engineering designs. Typical applications where interface friction values are essential include the design of piles, retaining walls, and more recently, contaminant storage ponds and heap leach pads lined with geosynthetics. An upper bound solution to the interface problem would assume the interface friction angle (8) equal to the internal friction (<])) of the soil in contact with the structural .surface. However, in most applications 8 is lower than < > | and will therefore be one of the governing factors in geotechnical design where the interface represents a potential failure surface. Often 8 values are specified arbitrarily for various contacting surfaces. A rational assessment of the mechanics and factors associated with interface shear considering mobilization of both peak (8p) and residual (Sr) states will clearly be of both fundamental and practical importance and may allow for less conservative 8 values to be applied in design. Interface friction parameters are usually obtained by one or a combination of the following methods: o Back analyses of failures of full scale structures; o Instrumented field experiments; and o Laboratory testing using conventional techniques modified for interface studies. Laboratory testing using a modified direct shear test is the most frequently used method for determining 8 values. However, progressive failure across the sample tested in direct shear may result in a measured 8 value considerably below the true peak. In addition, the  2 use of multiple reversals to attain residual values does not simulate field conditions where large relative displacements occur without changes in direction. Based on past findings, it has also become common practice to assign a 8 value which is an arbitrary fraction of ()) for surrounding the soil. These values are most commonly suggested for steel and concrete interfaces due to their historical applications and may range from as little as 20% to 100% of the internal friction angle of the surrounding soil. Recommended 8 values for design published in literature are often contradictory and generally fail to adequately address the factors influencing the interface friction angle.  Current literature pertaining to geosynthetic design recognizes that a  rational assessment of 8 is required.  Martin and Koerner (1985) present design  considerations for geomembrane lined slopes and express the need for an extensive data base of soil-geomembrane 8 values that represent the candidate materials involved. Laboratory studies in which the factors influencing interface friction are clearly identified will ultimately provide an aid to design and limit the future use of gross estimates of 8. Except in a limited number of studies, researchers have not addressed the manner in which the properties of the interfacing constituents may affect the observed 8 values. In addition, assessment of possible plastic deformation of the structural surface as a result of shearing has not been adequately investigated. Any assessment of 8 values must account for the properties of the contacting surfaces and relate these to the material behavior during relative motion.  This requirement necessitates the application of  traditional concepts in soil mechanics with those associated with the field of Tribology. Tribology is defined as "the science and technology of interacting surfaces in relative motion and of related subjects and practices" (Hailing, 1975). This thesis focuses on a laboratory study to evaluate the friction angle mobilized at the interface between sand and steel, concrete and geosynthetic surfaces . The investigations are carried out in a ring shear apparatus. The annular sample used in ring  shear testing minimizes the effects of progressive failure prevalent in direct shear tests and therefore allows for a better assessment of 5p. When combined with its capacity to realize 8f through unlimited relative displacement between the sand and the structural surface, theringshear apparatus is very desirable for interface testing. A systematic laboratory study of interface friction is presented. Two common quartz sands were used which have identical gradation and mean grain size. The grain shape of the two sands differs significantly with one sand being classified as round and the other angular. The structural surfaces used in this research program include smooth and rough mild steel, smooth mortar finished concrete and two varieties of geosynthetics classified as geomembranes. The two geomembranes used cover a range of soft and hard, and smooth and rough surfaces. A rigorous examination of surface roughness parameters is included in this study to assist with the interpretation of the observed interface friction values.  4  2. INTERFACE FRICTION - LITERATURE REVIEW 2.1 General The majority of reported interface friction values between sands and construction materials have been obtained using an assortment of modified direct shear tests. Although these data have been used to define both peak (5p) and residual (5r) interface friction angles, the extreme non uniformity of the strains across a sample tested in clirect shear, will result in a measured peak value below that which would exist in field conditions. Furthermore, the use of the direct shear test requires multiple reversals to observe residual values. This technique does not simulate field conditions where large relative displacements occur without changes in direction. In a series of stress and strain controlled tests on wood, steel, and concrete, Potyondy (1961) used both 60 and 90 mm shear boxes.  In studies involving  geosynthetics, Martin et al. (1984) used a 100 mm shear box, while an even larger box measuring 280 mm was used by Saxena and Wong (1984).  A typical testing  configuration using the modified direct shear test is illustrated in Figure 2.1.  For  relatively flexible surfaces such as geosynthetics, the material is usually attached to a rigid block. However, placement of the membrane on a sand bedding has also been used successfully (Eigenbrod and Locker, 1987). To overcome serious disadvantages of the direct shear apparatus Yoshimi and Kishida (1981) used a ring torsion apparatus for measuring interface friction parameters. In their study three types of sands were sheared using mild steel as the interface material. In 1989 Negussey et al. used a ring shear apparatus developed at the University of British Columbia (UBC) to evaluate the interface friction between granular materials  5  NORAAI—  STRESS  ///////  <l> D E F L E C T  SHEAR  °9LO°O  S O I L  S T R U C T U R A L  //w  °°QLO  IO N  yy \ \  S U R F A C E  //\\ //\\ // w // \\ //\\ // \\ // \  Figure 2.1  Typical Modified Direct Shear Apparatus For Interface Testing  STRESS  and geosynthetics. This apparatus appears to be much simpler to operate than that used by Yoshimi and Kishida (1981). Use of a simple shear apparatus for assessing interface friction between sand and steel is presented by Uesugi and Kishida (1986). This testing apparatus allows for the determination of interface friction parameters and has the capability to separate the deformations due to shearing strains in the sand from those due to sliding displacement at the interface. The majority of studies on interface friction do no adequately address in a controlled manner the effects of each of the factors among the many factors affecting observed 8 values. Often the properties of the structural surface or the interfacing soil are poorly documented. The following sections attempt to summarize some of the reported interface friction studies conducted in the laboratory in which sand has been used with steel, concrete, and geosynthetics.  2.2 Sand-Steel Interface Testing Data published by Potyondy (1961) using the modified direct shear test indicates interface friction angles ranging from approximately 240 to 34o for smooth and rough steels respectively. Surface roughness parameters associated with the terms "smooth" and "rough" are not provided. Negligible effects were found on the value of 8 as a result of increasing normal load (48 - 143 kPa) or shearing with dry or saturated sand. Only data for dense sands are presented and no discussion on the possible effects of density of the sand is provided. The type of sand (mineralogy) used for this study is not discussed; however, the angle of internal friction ranged from 390 to 440. An assessment of large strain interface values does not appear to have been addressed in this study.  7 Butterfield and Andrawes (1972) used a conventional 60 mm shear box to determine static (5p) and kinetic (8r) interface friction values for polished mild steel and Leighton Buzzard (B.S.S. 14-36) sand under low normal stresses (14 - 97 kPa). They observed 5p to range from 11.30 to 18o for loose and dense sands respectively while cV was shown to range from 9.8o to 15.6o. These observations indicate that both 8 p and 5r are a function of the sand porosity. This finding does not appear to be addressed in other studies on sand-steel interface testing. A surface roughness parameter associated with the term "polished steel" is not provided. Yoshimi and Kishida (1981) used the ring torsion apparatus to evaluate the friction between three types of sand and metal surfaces. They concluded that the range of mobilized friction angles for a given sand as a function of surface roughness was wider than previously reported. Smooth steel (R ma x = 2.3 um)l was shown to mobilize 5 p of approximately 12o and 8r of 10o while rough steel (Rmax = 500 pm) indicates 8 p = 37o and 8r = 36o.  For the stress range considered (51 - 158 kPa) 8 was found to be  independent of normal stress and the type (rnineralogy) of sand used. Data from this study were shown to correlate well with previously published data by Esashi et al. (1966) and Brummund and Leonards (1973) for smooth steel surfaces. Uesugi and Kishida (1986) present the results of a well documented and detailed testing program in which the simple shear apparatus was used to study factors affecting friction between steel and dry sands. They conclude that the type of sand (mineralogy and grain shape) and the surface roughness have a significant influence on the frictional coefficient while the grain size and normal stress are not so influential. In this paper a review of the ring torsion data from Yoshimi et al. (1981) is provided and a fundamental difference in sample preparation techniques is identified. This difference may in part explain why Yoshimi did not observe a significant effect of sand type on the frictional 1 Rmax is a surface roughness parameter defined in Section 2.5  coefficient or attain a frictional coefficient upper-bounded by the shear resistance of the sand for rough surfaces. Sand-steel interface friction parameters are quoted in a wide variety of reference texts such as that by Sowers and Sowers (1951). Here a range of 1 lo to 170 is suggested; however, as is common in many of these texts, no reference is made as to the origin of these values nor their relationship to the characteristics of the interfacing surfaces.  2.3 Sand-Concrete Interface Testing In the same paper discussed in Section 2.2 Potyondy (1961) observed 8 for smooth concrete formed against wood of approximately 390 while for rough concrete (j>opne8 a-yawo-i ooiX 8 increased to 43°. These data indicate that between 90% to 100% of the internal friction angle of the interfacing sand was mobilized at the small relative displacements in the direct shear test. The commonly used value of 8 = 2/3<J> is referenced in "Soil Mechanics in Engineering Practice" by Terzaghi and Peck (1948) in a discussion of concrete retaining walls. This value is frequently referred to in engineering practice although its origin is seldom recognized. Also present in a wide variety of reference texts such as Sowers and Sowers (1951) are typical tan 8 ranges of from 0.5 to 0.8 that have been used for design purposes.  2.4 Sand-Geomembrane Interface Testing The recent explosive growth of geosynthetic use in the market place has resulted in several studies being published, pertaining to material properties and related field design parameters.  A very limited number of studies have addressed the soil-  geosynthetic interface and the factors influencing the frictional properties. Published  9  data in this area shows that only modified direct shear tests have been used for these studies. Since this thesis focuses solely on sand-geomembrane interfaces, a discussion of cohesive soils and geotextiles is not included. For this reason all references cited refer to sand-geomembrane findings although a broader scope of information may be present in these references. A commonly used reference for interface friction design stems from results published by Martin et al. (1984) in which a series of low stress (14 - 103 kPa) direct shear box tests were conducted on a variety of geosynthetics and granular materials. The four types of geomembranes used in this study, EPDM, PVC, CSPE and HDPE represent the most commonly used surfaces for a variety of field applications. Efficiency ratios, E (E = tan 8/tan $), are shown to range from .60 to .96 for the sands and surfaces used in this study. The authors comment that these values represent both peak and residual friction angles since in this study the technique of multiple reversal passes produced no noticeable difference between 8 p and 8r. Although (J> angles for the sands are presented no distinction has been made as to weather these represent peak or residual values. The constant volume friction angle, <|)cv» may be a preferable reference since it is unique for a given sand and therefore independent of D50, C u , D r etc. Negussey et al. (1988) used the UBC ring shear apparatus for an interface friction study on geosynthetics and granular materials. Data from this study indicated that peak and residual 8 values were mobilized at the interface of the hard HDPE geomembrane at all stress levels and that these 8 values increased with the angularity of the sand. Negussey et al. (1988) obtained 8p = 17.6o which is in good agreement with results of Martin et al. (1984) for the Ottawa sand-HDPE interface condition. However, in this same study Negussey reports 8r = 15o for similar stress conditions. This finding is  inconsistent with Martin's observations that Sp = 5r and may be due to the inability of the multiple reversal technique to adequately simulate large strain interface conditions. Saxena and Wong (1984) using a very large modified shear box (280 mm) measured  8 p and 8r values that differed by approximately 1.50 for Ottawa sand-HDPE  specimens tested at 200 kPa normal load. They also observed an increase in 8 of approximately 3o between tests conducted at 70 kPa and 200 kPa. Similar observations are made by Eigenbrod and Locker (1987) where they observed that 8 p increased with increasing normal stress; this increase was observed at normal stresses less than 50 kPa.  2.5 Surface Texture Assessment As previously mentioned, the surface roughness of materials tested using interface shear techniques are typically described using adjectives such as "smooth" or "rough". An accurate characterization of surface texture using established roughness parameters will aid greatly in the understanding of the mechanics of interface shear. The quantitative assessment of surface texture is most commonly obtained by profilometry methods using a fine diamond stylus which traverses the surface while its vertical movements are recorded. Since vertical asperities are typically much smaller than the horizontal distance along the surface being tested, a distortion of the recorded shape of the profile may lead to a physical misconception as to the general character of the actual surface profile. However, assuming that a representative series of traces are performed, the profilometer provides useful measurements of the various surface texture parameters necessary for their characterization. Surface texture refers to the locally limited deviations of a surface from the smooth ideal. It is comprised of both a roughness component of the finer irregularities and a waviness component consisting of the wider-spaced deviations (see Figure 2.2).  Figure 2.2  Components of Surface Texture (after Schaffer, 1988)  12 The introduction of a significant waviness component would counter the assumption that relative displacement occurs along a planer boundary and would undoubtedly introduce scale effects regarding specimen size and configuration in the interpretation of test data. Surface texture is quantified by parameters which relate to certain characteristics of the texture. These parameters can be classified into the following categories: o  Amplitude parameters are measures of the vertical displacements of the  profile. When referred to the roughness component of surface texture, they are typically designated as R parameters. o  Spacing parameters are measures of irregularity spacings along the  surface, irrespective of the amplitude of these irregularities. o  Hybrid parameters relate to both the amplitude and spacing of the surface  irregularities. Although some 200 different surface-texture parameters have been developed, only two amplitude parameters have been presented in this study to aid in understanding the interface condition during shear. R a is the universally recognized and most widely used international parameter of surface roughness. It is the arithmetic mean of the departures of the profile from the mean line and is determined as the mean result over several sampling lengths L (see Figure 2.3a). It is useful for detecting general variations in overall profile height It cannot however detect differences in spacing and its distribution or the presence or absence of infrequently occurring high peaks and deep valleys. Rmax is the maximum peak-to-valley height of the profile within the sampling length (see Figure 2.3b). This parameter was used by Yoshimi and Kishida (1981) based on a sampling length of 2.5 mm. Subsequently, Kishida and Uesugi (1987) and Uesugi and Kishida (1986) incorporated the particle diameter of the sand into Rmax by  R  a  " L fo^  iv(«)ld«  Figure 2.3  Figure 2.3a  Ra  Figure 2.3b  Rmax  Surface Roughness Amplitude Parameters  14  modifying the gauge length equal to the D 5 0 of the sand. This was done to better correlate findings in which the surface exhibited a significant waviness component. Since the surfaces tested in this thesis did not possess a significant waviness component, the modified R x parameter was not used. ma  2.6 Summary The majority of studies pertaining to interface friction values fail to provide a systematic approach addressing the factors affecting measured 8 values.  The  characteristics of interface constituents and the specimen configuration are often inadequately documented which makes comparison of data from several sources difficult. A standardized methodology for interface friction testing has not been adopted. Although modified direct shear tests have been most frequently used, varying sample sizes and testing configurations may be the cause of inconsistent or conflicting fmdings from different studies. Data from the majority of direct shear interface friction studies do not indicate distinct peak and residual 8 values which may be the result of extreme nonuniformity of stresses across the sample tested in direct shear and the failure of this apparatus to provide continuous uniform large shear displacements. This study addresses the need for a systematic detailed study to identify the factors affecting interface friction between sand and steel, concrete and geosynthetic surfaces.  The ring shear apparatus is a desirable tool for interface testing due to its  inherent capability to permit unlimited displacement along the interface and less progressive failure across the sample than a specimen in direct shear.  15  3. EXPERIMENTATION 3.1 Ring Shear Testing - General Several researchers have developed, ring shear or torsion type devices in an attempt to overcome some of the disadvantages of the more universally accepted direct shear test Bishop et al. (1971) present a comprehensive review of ring shear devices developed until that date. It is interesting to note that in all references cited, the ring shear apparatus was used exclusively for the determination of the residual strengths of cohesive soils. In this paper Bishop et al. illustrate the superiority of the ring shear test over the multiple reversal direct shear test for this purpose.  3.2 UBC Ring Shear Apparatus The UBC ring shear apparatus was designed by Bosdet (1980) for investigating the residual strength of clays. It was recently modified for testing the frictional characteristics of granular materials at large strains (Wijewickreme,1986) and further adapted by Negussey et al. (1989) to evaluate the interface friction between sand and geosynthetics. The present configuration of the UBC ring shear apparatus is illustrated in Figure 3.1a. The only modifications made to the apparatus for this thesis was the upgrading of the data acquisition system. The previous arrangement of using two strip chart recorders to monitor loads and displacements has been replaced with a P.C. based data acquisition system.  This results in a fully automated testing process capable of scanning and  recording data at a maximum rate of 3 Hz. Six channels of data are monitored during each test and are presented in real-time text and graphics display for the convenience of  SPECIAEN  CONFIGURATIONS  DETAILED CROSS-SECTIONS fceni_ATs> AIR wm.t  SECTION A RIBBED  PLATEN  i t  -  3 33 9a go  SAND STEEL OR CONCRETE  _n_  SHEAR  STUD  SECTION B RIBBED  (TV  Si  PLATEN  SAND  QEOAEnBRANE BASE  PLATEN  SHEAR  STUD  SECTION C RIBBED  PLATEN  —  51  SAND GEOAEABRANE SAND RIBBED  PLATEN  3.1b  3.1a  Figure  UBC Ring Shear Apparatus  1  PNEUAATIC  2  LOADING  3  UPPER  CONFINING  RINGS  4  LOWER  CONFINING  RINGS  5  ROTATING  6  STRAIN  7  A / D  B  IBA P C  PISTON  YOKE  A N DUPPER  TABLE  DRIVE  CONVERTER  RIBBED  PLATEN  17  the operator. Variables monitored include normal and shear displacements, normal stress acting on the plane of shear, torsional stress, and the mobilized 8 value. The apparatus accommodates an annular sample with inner and outer radii of 44.5 and 70.0 mm respectively. The height of the sample is variable but typically averages 15 to 20 mm in the two halves of the rings.  The normal load on the failure plane is  calculated by subtracting the side friction on the upper confining rings (recorded directly by the bottom load cell) from the total applied normal force (recorded by the upper load cell). A major advantage of the UBC ring shear apparatus is the ability to apply a wide range of normal loads through a pneumatically actuated piston. The system is capable of a maximum pressure of over 1 MPa. The the current resolution of the normal and torsional load cells however, limits the accuracy of data below approximately 100 kPa.  3.2.1 Testing Configurations To accommodate the clifferent types of structural surfaces tested in this thesis, various  specimen configurations were used (see Figure 3.1b Sections A, B and C).  Section A illustrates the configuration used for the sand-steel and sand-concrete interfaces. Here the prepared steel or concrete specimen is positioned in the annular space bounded by the inner and outer lower confining rings, such that the upper surface is coincident with the top of the rings. The specimen is held firmly in place by a series of shear studs attached to the base of the rotating table. After placement of the inner and outer upper confiningrings,the sand specimen was prepared by air pluviating dry sand in a loose condition. The surface of the sand was levelled by vacuum suction before the upperribbedplaten and loading yoke were positioned.  18  Figure 3.1b Section B shows one of two configurations used for the sandgeomembrane interface. In this configuration the geomembrane was glued to a smooth rigid base platen such that the interface was coincident with the top of the lower rings. Base platens of various thicknesses were necessary to accommodate the range of geomembrane thicknesses tested. Section C (fig 31.b) represents the alternative method used for assessing sandgeomembrane friction. In the area bounded by the inner and outer lower confining rings, a thin ribbed base platen was firmly secured to the base. The sand was air pluviated above and levelled with the top of the rings. The pre-cut geomembrane specimen was then placed on the sand surface in such a manner as to minimize any disturbance. This was followed by air pluviating dry sand, levelling, and finally placement of the upper ribbed platen and loading yoke.  3.3 Test Materials 3.3.1 Sands Two quartz sands with essentially identical gradation and median grain size but different grain shapes were used for this study (see Figure 3.2). This enabled the effect of grain shape alone on 8 to be isolated from other characteristics of the sand. Medium Ottawa C-109 sand is a uniformly graded quartz sand with rounded particles and has a median grain size of 0.4 mm. Target 20-30 sand differs significantly from Ottawa sand only in its particle shape which is described as angular. This sand originates from the state of Washington and its angularity is the result of mechanical crushing. The photomicrograph shown in Figure 3.3 illustrates the marked contrast in angularity of the two sands.  19  I I  T  I  0  0.01 GRAIN SIZE, mm  Figure 3.2  Ottawa and Target Sand Grainsize Distribution  Figure 3.3  Photomicrograph of Ottawa and Target Sand Grains  21  Prior to the start of the using the  interface testing program a series of tests were conducted  UBC ring shear apparatus to determine the constant volume friction angle,  of Ottawa and Target sands. The has  <(>cv =  (j>Cv>  results of this series of tests indicate that Ottawa sand  290 and is essentially independent of the  normal stress applied on  shear. Target sand however exhibits an average §  cw  the plane of  = 33o over the range of stresses  applied. For this angular sand significant particle crushing occurred at normal stresses above 100 kPa.  <t>Cv is used as the reference friction parameter of the  soil because it is a  constant for a given sand and unlike <|>peak is independent of packing density, gradation and  normal stress. For both Ottawa and Target sand pluviated in a loose condition, the  internal friction angle (<{>) at small strains is essentially equal to the constant volume friction angle  ((]>cv)-  3.3.2 Construction Materials The  structural surfaces chosen for interface testing with Ottawa and Target sand  include steel, concrete, and two types of geomembranes. While steel, concrete (and wood) have historically been the  construction materials used in geotechnical practice, the  increasing use of geosynthetics has created a need to obtain representative design parameters, including interface friction values. In this study detailed information regarding the preparation of surface roughness and  testing of the structural surfaces is included, thus avoiding many of the omissions  identified in the  papers reviewed in Section 2.  3.3.2.1 Steel Specimens were cut from a length of seamless mechanical tubing, SAE 1015 low carbon steel. Each specimen was first parted from the tubing section and the top and  22  bottom surfaces ground flat and parallel. Eight equally spaced holes 4 mm deep were drilled into the bottom of the specimen to match the shear studs affixed to the base platen in the testing apparatus. The upper surface was then prepared for interface testing using a series of multiple straight line passes on a surface grinder. This process produced an average R a of approximately 0.3 um (R ax = 3 urn). This surface roughness is similar to m  that for smooth steel tested by Yoshimi and Kishida (1981) which was created by a technique of lapping on fine emmery cloth with spindle oil. As a result of the grinding process, anisotropic surface roughness is imparted to the specimen. Measurements taken in the direction of the grinding stone travel indicate a typical R a of 0.2 um while at right angles to the travel Ra = 0.35 um. This roughness anisotropy was not considered significant and was therefore ignored when presenting surface roughness data. The surfaces addressed in this study were prepared and tested in the ring shear apparatus in such a way as to minimize the waviness component Rough surfaces were prepared by sand blasting a smooth surface specimen using fine glass beads. This technique produced a uniform R a across the sample of approximately 2.5 um (Rmax = 20 um). It is important to note that this process does not alter the surface waviness of the specimen.  Surfaces rougher than R a = 2.5 um were not  tested in this study. Prior to placement in the ring shear apparatus the surface of each specimen was cleaned with a degreasing agent to minimize any possible lubrication effects that might result from the presence of residual fluids from the grinding process.  3.3.2.2 Concrete Concrete specimens were prepared using ASTM C109 Standard Test Method for Compressive Strength of Hydraulic Cement Mortars. The mortar composition consisted  23 of one part cement to 2.75 parts medium Ottawa sand with a water-cement ratio of 0.485. The samples were formed in a two part plexiglass mould and allowed to cure in a saline bath at 100% humidity for not less that 28 days. This standard procedure minimizes the effects of time dependent strength increase of concrete. After curing, the surface of each specimen was cleaned with fresh water and allowed to dry for 24 hours to minimize any possible effects that might result from the presence of residual deposits while the specimen was in the water bath. Use of the profilometer for surface texture assessment was not applicable for the concrete surfaces due to significant non-uniformity of the surface and size of the asperities. The problems associated with the forming of the annular specimens inhibited the ability to produce consistent and reproducible surface textures for all but the "smooth" surface.  To attain a smooth surface the specimen was vibrated during  placement of the mortar and covered with a plexiglass plate during the initial phase of curing.  3.3.2.3 Geomembranes Of the major types of geomembranes in current use, two types were selected for testing in this research program. Polyvinyl chloride (PVC) is characterized as being medium stiff with a soft surface, rough on one side with R a = 2.3 um (Rmax = 17 um) and smooth on the other (Ra = 1.1 um, R m a x = 9 um). Note, the absolute value of these roughness parameters may be inaccurate due to the soft nature of the surface. However, the presence of smooth and rough sides was noted from visual observations. Specimens 20 and 30 mils thick were tested in this study. In contrast to PVC, High Density Polyethylene (HDPE) is stiff and hard and is manufactured with both sides smooth or roughened. Smooth HDPE specimens 20 and  24  100 mils thick were tested. These specimens are characterized by a typical  value of  1.3 um (R ax = 8 um). Roughened HDPE is available only in 100 mil thickness and m  could not be tested with the profilometer due to the large size of the asperities. Based on visual observations the asperities are distributed randomly across the surface and extend approximately 0.5 mm above the surface. All geomembrane specimens were trimmed to annular shape for testing in the ring shear apparatus by cutting with a carbon steel cutter. Prior to placement in the ring shear apparatus the surface of each specimen was cleaned with a degreasing agent to minimize any possible lubrication effects that might result from the presence of a contaminating film.  3.4 Testing Program To represent the low overburden pressures acting on the geosynthetic interface of pond liner systems, researchers have conducted tests typically ranging from as litde as 14 kPa (Martin et al., 1984) to as high as 200 kPa (Saxena and Wong, 1984). Based on this range and the optimum measurement resolution in the present configuration of the UBC apparatus, low stress tests were confined to 100 kPa normal stress acting on the plane of shear. In order to simulate the much higher stresses that would be acting on a geosynthetic liner beneath a heap leach pad where the fill can be several tens of meters high, a maximum stress of 750 kPa was chosen to represent the high stress condition. Although the UBC ring shear apparatus is capable of applying more than 1 MPa normal stress, problems associated with specimen deterioration at these high stresses hampers confident test interpretations. Intermediate tests at 350 and 500 kPa were also performed  25 to detenrrine possible effects of the normal stress on the mobilized friction angle at the soil-geosynthetic interface. Following an initial series of tests, the dependence of 8 on the magnitude of relative shear displacement between the sand and the geosynthetic became evident. Although peak (small strain) friction values are of prime importance in geosynthetic design it was decided to conduct all tests to a shear displacement of 350 mm in order to observe the large strain response. The same testing schedule as described above for geomembranes was used for the steel and concrete specimens. However, tests were continued to 650 mm in an attempt to observe the large strain response. This represents more appropriately the insertion of a pile or cone penetrometer into sand where large differential displacements occur at the soil-wall interface. In order to rmnimize possible effects of varying sand density on 8, all tests were conducted with sand pluviated in air at similar relative densities. For convenience, sand in a loose condition was pluviated with relative densities not exceeding 25%. In this loose state the internal friction angle (<()) at small strains is essentially equal to the constant volume friction angle (§ C \)All horizontal displacements referred to in this thesis correspond to those at the mean radius of the annular specimen. Although displacement rates for the UBC ring shear apparatus can be varied from 0.01 mm/hour to 1000 mm/hour, those used in this research program averaged 300 to 500 mm/hour. Since displacement rate does not influence 8 (Negussey et al., 1988), this rate was selected for convenience of data acquisition. This when coupled with a scanning frequency of 3 to .2 Hz provided sufficient detail to accurately monitor instrumentation response while allowing for manageable data files for data reduction and presentation.  26 A Talysurf 5 profilometer coupled with an IBM P.C. were used to determine the roughness parameters for all structural surfaces tested in this research program. Roughness parameters R a and R x were obtained from a series of traverses (typically 4 ma  to 6) extending radially across the specimen. The average value from these traces was then assigned as the representative value for the surface. Surface roughness measurements for each specimen were taken before and after testing in the ring shear apparatus in order to correlate the effect of surface scour with the structural surface, grain shape of the sand and confining pressure.  27  4. GEOMEMBRANE INTERFACE TEST RESULTS Many different types of geomembranes have been developed and are primarily used for providing an impermeable boundary to fluid flow. Polyvinyl Chloride (PVC) and High Density Polyethylene (HDPE) are two of the most commonly used geomembranes that have been tested in this thesis. They represent relatively soft and hard surfaces respectively.  4.1 PVC Typical results relating 8 with horizontal shearing displacement for the Ottawa sand-PVC interface are shown in Figure 4.1. Tests conducted at 100 and 500 kPa normal stress are presented for shear displacement from 0 to 25 mm. The (J)Cv for Ottawa sand of 290 is also shown in this figure. These profiles clearly indicate that at both 100 and 500 kPa 8 is essentially equal to <j)cv and that 8p is mobilized by approximately 2 - 3 mm relative displacement along the interface. With increasing shear displacement 8 values show no appreciable reduction from the peak condition and thus 8r is approximately equal to Sp. The friction mobilized at the Target sand-PVC interface is illustrated in Figure 4.2. These curves also indicate that 8 is approximately equal to (t>cv OfevTarget= 33o) at both high and low stresses; however, 8V is some 2 to 3o below 8p. Visual inspection of the PVC specimens after testing indicated no continuous circumferential scour grooves on the surface of the PVC for either sand at any stress level. This finding is similar to that noted by Negussey et al. (1988) for HDPE. Under microscopic examination the surfaces tested at high normal stresses showed occasional  28  Figure 4.1  Interface Friction vs. Shear Displacement Ottawa Sand - PVC Interface (0 - 25 mm)  Figure 4.2  Interface Friction vs. Shear Displacement Target Sand - PVC Interface (0 - 25 mm)  pits or pockmarks and random scratches.  However, continuous grooves in the PVC  surface was not observed. This visual observation is consistent with the findings that both sands exhibit 8 values approximately equal to <J)CV implying that relative displacement is occurring within the sand rather than along the sand-PVC interface. If significant relative displacement was occurring along the sand-PVC interface, regional plastic deformation of the PVC would be expected. These findings are supported by the surface roughness data presented in Figure 4.3 which indicates that after testing in the ring shear apparatus over a range of normal stresses, R a of the PVC is virtually unchanged from that before testing. No significant difference in measured 8 values with geomembrane thickness was observed in tests using either Ottawa or Target sand in which 20 and 30 mil PVC geomembranes were used. Also specimen configuration was found to have little effect on 8. Tests conducted with the PVC specimen affixed to a rigid base (Section B Figure 3.1b) showed essentially no difference in 8 during the initial stages of shear when compared with tests conducted in which the specimen was positioned with a sand interface on both surfaces (Section C Figure 3.1b). Typical data illustrating this response are shown in Figure 4.4. It should be noted however, that Section C type specimens did show slightly increasing 8 values as relative displacement was carried out to very high values. This can be attributed to undulations or folds in this soft PVC geomembrane that developed during these tests. Since folds could not develop during testing of Section B type specimens, 8 values in these tests remained constant at large relative displacements. The coefficient of friction | i (|i = tan8), when plotted vs. normal stress is a convenient means of presenting the data for all the tests conducted in this research program. This allows for trends in the data to be identified and facilitates comparison with values presented by other researchers.  Figure 4.5 illustrates the relationship  between the coefficient of friction at the peak condition (p:p) and the normal stress at  31  10.0  200  3 0 04 0 05 0 06 0 0  NORMAL STRESS ON PLANE OF SHEAR, k P a  Figure 4.3  Roughness Coefficient vs. Normal Stress After Shearing Ottawa Sand - PVC Interface  7 0 8 00 0  Figure 4.4  Interface Friction vs. Shear Displacement Ottawa Sand - PVC Interface (0 - 25 mm)  33  1.0 0.9 * 0.8 go 0 . 7  UTARGETSAND  1^  K  ^ 0.6 o 0.5  A A  .A A  ^OTTAWA SAND  E-H  2; w  0.4 1 0.3 0.2 -  * * * * * OTTAWA SAND AAAAA TARGET SAND  0  0.1 0.0  0  "i  1—  100  200  3 0 04 0 05 0 06 0 0 7 0 0 T  T  NORMAL STRESS ON P L A N E OF SHEAR, k P a  Figure 4.5  Interface Friction vs. Normal Stress Ottawa Sand, Target Sand - PVC Interface  800  which Ottawa sand-PVC and Target sand-PVC tests were carried out. Between 100 and 500 kPa, Hp may be seen to be approximately equal to 0.55 for the Ottawa sand-PVC interface and 0.65 for the Target sand-PVC interface. These values indicate that an interface friction equal to the full strength of the soil was being mobilized regardless of the normal stress for both sands.  These data also suggest that there is a trend of  increasing u p at higher normal stresses, however it does not appear to be significant. Values of u from Martin et. al (1984) for PVC interfaces with concrete sand (<)> = 30o) and mica schist sand (<() = 26°) have been included for reference in Figure 4.5. These data plot slighdy below the Up values found in this study; however, they do indicate reasonable agreement.  The mineralogy for these sands are not specified  although it is reasonable to assume that the concrete sand is composed primarily of quartz. Unfortunately specific data for the Ottawa sand-PVC interface are not presented by Martin et al. (1984).  4.2 Smooth HDPE Typical profiles showing 8 as a function of shearing displacement for the Ottawa sand-HDPE interface are shown in Figure 4.6. Tests conducted at 100 and 750 kPa normal stress are presented for horizontal displacements from 0 to 25 mm. Although these tests where continued for a total of 350 mm no change in 8 values occurred past about 25 mm. For comparison with the angle of shearing resistance of the sand alone where Ottawa= ^9°, Figure 4.6 clearly indicates that 8 which ranges from approximately 14 to 210 is substantially below § C \- This figure also reveals that there are distinct peak and residual interface friction components that are dependent on the amount of relative displacement and the value of the normal stress. 8p shows an increase from 18 to 210 as a result of an increase in normal stress from 100 to 750 kPa. The amount of relative  35  Figure 4.6  Interface Friction vs. Shear Displacement Ottawa Sand - HDPE Interface (0 - 25 mm)  36 displacement required to mobilize this peak value is approximately 1 mm at both stress levels. After attainment of 5p the 8 variations at both the high and low stress have similar shapes whereas, 8r at 100 kPa appears to reach a limiting value of about 14o by 5 mm displacement, at 750 kPa the Umiting oV of about 170 requires much larger displaceemnt of approximately 25 mm. Initial examination of Figure 4.7a suggests that for the Target sand-HDPE interface 5 values of about 2 7 ° to 30° are mobilized. The peak is reach at displacements less that about 5 mm and residual values are approximately equal to the peak up to about 25 mm shear displacement.  However, when the horizontal displacement scale is  expanded to include the full 350 mm displacement to which the tests were conducted, it becomes evident in Figure 4.7b that with large shear displacements significant changes in 8 occur. At low stresses 8 continues to reduce gradually and does not reach a limiting value by the end of the test at 350 mm. This is in stark contrast to the behavior at high stress that show a rapid decrease in 8 between approximately 30 mm and 75 mm before approaching a limiting value of about 2lo after a shear displacement of about 100 to 200 mm.  The surface roughness parameter R a , of HDPE specimens after testing is correlated with the normal stress in Figure 4.8. Tests with Ottawa sand show only a nominal increase in R a between 100 and 750 kPa while Target sand shows very significant increase. This indicates that a greater amount of plastic deformation occurs on the surface of the HDPE for a given normal stress when angular Target sand is used as compared to the rounded Ottawa sand. Typical roughness profiles presented in Figure 4.9 illustrate the effects  of  particle shape on the measured R a values. For tests conducted at 500 kPa normal stress, scouring is more pronounced for the HDPE-Target sand interface than the HDPE-Ottawa  40 3 5 -^-<t>cv ^30  CO  \-  \^  \y  / / / / /J // // I/ / f 11  §25 in £20  -  II II it  O  £ 15 W E-< 5  10  100 kPa 750 kPa  5H 0  0  i  i  i  i  1  i  i  i  i  1 10  i  i  i  1i 51  i  i  i  SHEAR DISPLACEMENT ( m m )  Figure 4.7a  i  1  20  Interface Friction vs. Shear Displacement Target Sand - HDPE Interface (0 - 25 mm)  i  i  i  i  25  38  40  Figure 4.7b  Interface Friction vs. Shear Displacement Target Sand - HDPE Interface (0 - 350 mm)  39  10.0  t±i±_* OTTAWA SAND AAAAA TARGET SAND  9.0 8.0 7.0 6.0 5.0  AA  4.0 3.0  A „--' A'  2.0 1.0 - R 0.0 0  a  before shearing  1 0 0 2 0 03 0 04 0 05 0 06 0 0  NORMAL STRESS ON P L A N E OF SHEAR, k P a  Figure 4.8  Interface Friction vs. Normal Stress Ottawa Sand, Target Sand - HDPE Interface  7 08 000  40  v  BEFORE TESTING  Smooth HDPE R a = 1.3 Um  AFTER TESTING @ 500 kPa NORMAL STRESS  R m a x = 5 Urn  Ottawa sand-Smooth HDPE Ra = 2.2 M m  Rmax = 22Ura  Target sand-Smooth HDPE Ra = 4.2um  1 T  250 M>"  -1 i*~  Figure 4.9  Typical Surface Roughness Profiles Ottawa Sand, Target Sand - HDPE Interface  R m a x = 34u m  41  sand interface. This qualitative assessment is confirmed by the measurements of R a for these interfaces which equal 4.2 um and 2.2 um respectively (fig 4.8). The plastic deformation associated with interface shearing in the ring shear apparatus appears as a series of concentric circumferential scour grooves as shown in the photomicrograph in Figure 4.10. Scour grooves of this type were observed in all tests conducted in this research program. For a given stress level grooves in the HDPE geomembrane were always larger for the angular Target sand than the more rounded Ottawa sand interface. It is important to note that while the grooves in Figure 4.10 appear quite large, the maximum depth scour (R m a x ) typically ranges from 20 to 30 um. This represents less that 10 % of the D50 for the sand grains. It follows therefore, that only the tips of the sand grains are in intimate contact with the HDPE.  Due to the  vertical exaggeration required to observe the surface roughness using the profiling gauge it is difficult if not impossible to identify individual scour grooves. To better understand the mechanics associated with the mobilization of 6p and 6V at the sand-HDPE interface, a series of tests was conducted at a constant normal stress of 350 kPa and terminated at different values of horizontal displacement. The results shown in Figure 4.11 illustrate that the horizontal displacement at which full development of scour (represented by a limiting value of Ra) is realized is coincident with the displacement at which the initial attainment of a limiting value of oV is reached. For the HDPE-Ottawa sand interface, Figure 4.6 showed that after approximately 10 mm shear displacement essentially no change in Sr occurs which is reflected in the R a parameter here in Figure 4.11. Similarly for the HDPE-Target sand interface, Figure 4.7b showed 8 is essentially unchanging beyond about 100 mm shear displacement which is also reflected in R a in Figure 4.11. These findings indicate that at the point the scour grooves are fully developed an equilibrium state is attained following which no further change in measured interface friction angle occurs.  Figure 4.10  Photomicrograph of Grooves Ploughed in HDPE  43  10.0 9.0  OTTAWA SAND " " 4 TARGET SAND  8.0  a' = 350 kPa  7.0 6.0 5.0 4.0 -  ^•limiting 8r  3.0  A,  2.0  limiting 8r  1.0 0.0  Ra before shearing —I  0  50  1  1  1  1  1  1  1  1 0 0 1 5 0 2 0 0  Figure 4.11  1—  SHEAR DISPLACEMENT ( m m )  2 5 3 00 0  Roughness Coefficient vs. Normal Stress Ottawa Sand, Target Sand - HDPE Interface  350  An explanation for the development of peak and residual friction angles mobilized at the HDPE-sand interface may be advanced from the Theory of Adhesion Friction after Bowden and Tabor (1964). In this theory the total friction force is the sum of the force necessary to cause plastic flow (adhesion term) plus the force required to plough hard asperities through a softer surface (ploughing term). For most metal surfaces this ploughing term denoted by p e is negligible in comparison with the adhesion term, but it may be important when relatively hard asperities slide over a softer surface (see Figure 4.12). This situation closely models that of a sand grain shearing along the surface of HDPE.  This simple model shows that p e is a function of yield pressure of the flat  surface, the number of asperities and the geometry of the asperity. It would be expected therefore that the angular Target sand with its sharp edges (smaller 0) would exhibit a larger ploughing term than Ottawa sand resulting in higher 5p values. When the observations relating the development of scour grooves from Figure 4.11 for the Ottawa sand-HDPE interface are considered along with the ploughing effects discussed above, it does not seem unreasonable to assume that the ploughing term is only significant during the initial stages of shear while the grooves are being formed. Therefore, the mobilization of distinct 8 p and 8r values would be expected as the ploughing term diminishes and has negligible effect after the scour grooves are fully developed. Figure 4.13 illustrates the relationship between the coefficient of friction at the peak condition (np) and the range of normal pressures at which Ottawa sand-HDPE tests were carried out. Between a stress level of 100 and 750 kPa, (i p shows a slightly increasing trend with stress and ranges from approximately 0.35 to 0.4. This rise in | i p may be attributed to deeper scour grooves in the surface of the HDPE with increasing stress. Also plotted on this figure are 8 p values from Martin et al. (1984) and Saxena and  45  RELATIVE AOTION E>  F = nrhpo where; F = friction force n = number of asperities r = embedded radius h = depth of embedment p0= yield pressure of soft surface  Figure 4.12 Hard Asperity Ploughing Through Soft Surface (after Hailing, 1975)  46  1.0 0.9 5-0.8 | 0.7 o i  ^OTTAWA SAND  OH  ^ 0.6 &H  o  E-H  $  0.5  Saxena £ Won ; wonj;  K&rtin et al.  w o 0.4 § 00. .13  00 .. 02 0  1 0 02 0 03 0 04 0 05 0 06 0 0  NORMAL STRESS ON PLANE OF SHEAR, k P a  Figure 4.13  Peak Interface Friction vs. Normal Stress Ottawa Sand - HDPE Interface  7 0 8 00 0  Wong (1984) for Ottawa sand-HDPE tests. These data show good agreement with the results obtained in this study. u r values for the Ottawa sand-HDPE interface shown in Figure 4.14 range from 0.28 to 0.32 and again show a somewhat increasing trend with stress. The u r value at 200 kPa normal stress interpreted from Saxena and Wong (1984) has also been included on this figure. This value of 0.36 is well above the residual value observedfromtests in this study. No explanation is forwarded for this apparent discrepancy other than the observation that the direct shear apparatus used by Saxena and Wong (1984) measures 280 mm which is by far the largest size shear box used for interface friction testing. Figure 4.15 illustrates the relationship between u p and normal stress for the Target sand-HDPE interface. As with the Ottawa sand-HDPE interface, u p shows a slighdy increasing trend with stress. For the range of stresses considered u p ranges from approximately 0.55 to 0.65, which is considerably larger than for the Ottawa sand-HDPE interface. This clearly indicates that higher friction values are realized at the sand-HDPE geomembrane interface as a result of shearing with a more angular sand. The residual interface friction angle for Target sand-HDPE interface as a function of relative displacement has already been shown to be much more dependent on normal stress than the Ottawa sand-HDPE interface. Figure 4.16 illustrates this relationship by plotting u r at several shear displacements for a series of tests conducted between 100 and 750 kPa. At a normal stress of 100 kPa, u decreases at a relatively constant rate with  48  1.0 0.9 3 0.8 |o 0 . 7 0.6  HOTTAWA SAND  o 0.5 W o 0.4 § 0.3 E-i  Saxena t Wong  i—i  0.2 H o.i  0.0  H 0  100  T  200  3T 0 0 4T 0 0 5 0 0 6 0 0  NORMAL STRESS ON PLANE OF SHEAR, k P a  Figure 4.14  Residual Interface Friction vs. Normal Stress Ottawa Sand - HDPE Interface  7 0 8 00 0  49  200  3 0 04 0 05 0 06 0 0  NORMAL STRESS ON PLANE OF SHEAR, k P a  Figure 4.15  Peak Interface Friction vs. Normal Stress Target Sand - HDPE Interface  7 0 8 00 0  1.0 0.9 it 0.8 2  NORMAL STRESS ON PLANE OF SHEAR, k P a  Figure 4.16  Residual Interface Friction vs. Normal Stress Target Sand - HDPE Interface  shear displacement throughout the test. As the normal stress increases so does the rate at which an ultimate limiting value of [L is reached.  4.3 Rough HDPE In the previous section the HDPE specimens tested could be described as smooth, with an R m a x value prior to shearing of about 5 um. Rough HDPE on the other hand is characterized by a textured surface of random asperities. The significant height of these asperities (about 500 um) precludes using the profilometer for characterizing the surface roughness. Typical results relating 5 with horizontal shearing displacement for the Ottawa sand-rough HDPE interface and the Target-rough HDPE interface are shown in Figures 4.17 and 4.18 respectively. Tests conducted at 100 and 500 kPa normal stress are presented for shear displacement from 0 to 25 mm. These figures clearly show that at both 100 and 500 kPa, 8 is essentially equal to (|>cv Visual observation of the rough HDPE specimens after testing showed that circumferential scour grooves did not develop and only minor abrasion of the asperities occurred.  This visual observation along with the 8 values presented above would  indicate that relative displacement was cccurring within the sand rather that along the sand-rough HDPE interface. Figure 4.19 illustrates the relationship between u p and the normal stress at which Ottawa-rough HDPE and the Target-rough HDPE tests were carried out. Between 100 and 500 kPa up was approximately equal to 0.55 for the Ottawa sand-rough HDPE interface and 0.65 for the Target sand-rough HDPE interface, indicating that the full strength of the soil was being mobilized for both sands.  Figure 4.17  Interface Friction vs. Shear Displacement Ottawa Sand - Rough HDPE Interface (0 - 25 mm)  Figure 4.18  Interface Friction vs. Shear Displacement Target Sand - Rough HDPE Interface (0 - 25 mm)  54  1.0 0.9  £0.8 g 0.7 o  11TARGET SAND  -A -*  OH  ^ 0.6 o 0.5 G 0.4  IX  E-  W3TTAWA SAND  * * * * * OTTAWA SAND AAAAA TARGET SAND  &H  I 0.3  0.2 0.1 0.0 0  — i  1  1  1  1  1  1  1  1  1  1  1  1  1 0 0 2 0 03 0 04 0 05 0 06 0 0  1—  NORMAL STRESS ON PLANE OF SHEAR, k P a  Figure 4.19  Peak Interface Friction vs. Normal Stress Ottawa Sand, Target Sand - Rough HDPE Interface  7 08 0 00  55  4.4 Comparison of Surface Performance To this point interface friction data has been presented separately for each of the geomembranes along with a discussion detailing  the response observed.  Since  ultimately the performance of various geomembranes must be compared, this section presents data for the small shear displacement or peak condition for the PVC, smooth HDPE and rough HDPE surfaces together, to illustrate their relative response to shearing with contacting Ottawa and Target sands. Figure 4.20 and 4.21 compares up values for the three geomembrane surfaces with Ottawa and Target sand interfaces respectively. For both sands the soft PVC and the much harder rough HDPE mobilize an interface friction that equals the full shear resistance of the sand while the smooth HDPE mobilizes u,p of approximately 65% for Ottawa sand and 90% for Target sand. These figures clearly indicate that for quartz sands Up is a function of the type of geomembrane, the surface texture, the angularity of the interfacing sand, and to a lesser extent the normal stress acting on the plane of shear. Based on the above observations any attempt to assign a single u value that would cover the types of sands and geomembranes tested in this study would be inappropriate and misleading. However, these data do suggest that rough HDPE provides significantly improved interface performance over smooth HDPE and essentially equals that of the much softer PVC. For the sand-smooth HDPE interface, the use of an angular sand will result in a higher interface friction value than if rounded sand is used.  56  W3TTAWA SAND  0.2 H \  0  OTTAWA-PVC INTERFACE OTTAWA-SMOOTH HDPE INTERFACE OTTAWA-ROUGH HDPE INTERFACE  1  1  I  I  I  I  1 0 0 2 0 03 0 0  I  |  400  I  1  1  1  1  5 0 06 0 0  NORMAL STRESS ON P L A N E OF SHEAR, k P a  Figure 4.20  Peak Interface Friction vs. Normal Stress Ottawa Sand - PVC, HDPE Interfaces  1  7 08 0 00  57  W A R G E T SAND  TARGET—PVC INTERFACE _ _ TRAGET-SMOOTH HDPE INTERFACE TARGET—ROUGH HDPE INTERFACE  0  1 0 0 2 0 03 0 0  400  5 0 06 0 0 7 0 08 0 0  NORMAL STRESS ON P L A N E OF SHEAR, k P a  Figure 4.21  Peak Interface Friction vs. Normal Stress Target Sand - PVC, HDPE Interfaces  5. STEEL INTERFACE TEST RESULTS Smooth and rough mild steel specimens were prepared using the procedure outlined in section 3.4.1. Extremely rough or rusted specimens were not tested because of the difficulty in reproducing identical roughnesses among tests.  5.1 Smooth Steel Figure 5.1 illustrates the friction angle mobilized at the interface of smooth mild steel (Ra=0.3 um) and Ottawa sand at normal stresses of 100, 500 and 750 kPa. This figure shows that at all stresses 8p = 16o and is mobilized within the first 1 mm of shear displacement. Continued displacement indicates that 5 decreases at essentially constant rate and by 25 mm shear displacement is approximately equal to 12o. Expanding the horizontal displacement scale to 650 mm, the curves from Figure 5.1 have been replotted in Figure 5.1b.  They show that at very large relative  displacements 8 values have a marked dependence on normal stress. At the low stress 100 kPa, the friction angle of about 12o remains essentially constant to a displacement of 650 mm. At higher stresses of 500 and 750 kPa the friction angle decreases rapidly after the peak from 160 to about 120 by 50 mm displacement. Beyond 50 mm a significant increase of 8 is observed and under 750 kPa normal stress 8 increases at a greater rate until it approaches a limiting value of about 27 to 28o. The response of smooth steel interfacing with Target sand at 100, 500 and 750 kPa normal stresses shows the same characteristic behavior as observed for the Ottawa sand-smooth steel interface at both small and large shear displacements (see Figures 5.2a and 5.2b). The peak friction angle of 240 changes only slightly with shear displacement  59  40 35 H to w  -30  cv  is  <  §25 E-<  S 20 w o <d  5  15H  10 HI 500 kPa 750 kPa 100 kPa  5^  0  0  ~i—i—i—i—|—r  5  ~i  i  i  10  i  i  i  \  i  15  i  i  i  SHEAR DISPLACEMENT ( m m )  Figure 5.1a  i  i  20  i—i—i—i—  Interface Friction vs. Shear Displacement Ottawa Sand - Smooth Steel Interface (0 - 25 mm)  25  60  40 to  35-  5-  500 kPa 750 kPa — 100 kPa  -  1  0  1  I  1  I  1  I  '  1  i  1  i  I  i  1  1  1  1  1  1  1  1  1  1  1  1  1  50 100 150 200 250 300 350 400 450 500 550 600 650 SHEAR DISPLACEMENT ( m m )  Figure 5.1b  Interface Friction vs. Shear Displacement Ottawa Sand - Smooth Steel Interface (0 - 650 mm)  61  cv  100 kPa 500 kPa 750 kPa  5-1  0  0  ~i  1  r  ~i  i  i  i  |  10  i  i  i  i  |  i—i—i—i—|—i—i—i—r~  1 5  SHEAR DISPLACEMENT ( m m )  Figure 5.2a  2 0  Interface Friction vs. Shear Displacement Target Sand - Smooth Steel Interface (0 - 25 mm)  2 5  Figure 5.2b  Interface Friction vs. Shear Displacement Target Sand - Smooth Steel Interface (0 - 650 mm)  63 at a normal stress of 100 kPa. At 500 and 750 kPa normal stress a decrease in 8 to about 2lo by approximately 50 mm followed by an increase with further shear displacement until a limiting value of approximately 290 is reached by about 300 mm. Visual inspection of the steel surfaces after testing indicates that the angular Target sand causes more wear than round Ottawa sand and that surface wear was more pronounced at higher stresses with each sand. This was evidenced by the presence of grey colored crushed sand grains within the shear zone adjacent to the steel surface. Scour grooves although present, were not as prominent as those observed on comparatively much softer HDPE geomembrane. This implies that ploughing of the steel surface by the sand grains was not a significant factor. These observations are substantiated in Figure 5.3 which shows generally low R a values for the steel surfaces for both sands after testing and a slight increase in R a with increasing stress. A series of tests over a range of horizontal displacements were not conducted as part of this research program. It is doubtful if the simple observation of a change in R a would help explain the complex relationship of 8 with increasing stress and displacement because of its relative insensitivity for the sand-steel interface. A more detailed study of wear rates in which the volume of material removed for a given displacement might be a more appropriate method of analysis.  5.2 Rough Steel Rough steel specimens characterized by R a = 2.5 urn exhibit similar trends in 8 with increasing stress and displacement when compared with smooth surfaces. As shown in Figure 5.4a, 8p values for the Ottawa sand-rough steel interface are approximately 220. At large shear displacements (Figure 5.4b) and high stresses of 500 and 750 kPa, 8  64  9.0  *±+±* OTTAWA SAND AAAAA TARGET SAND  8.0 H 7.0 6.0 5.0 H 4.0 3.0 H 2.0 -A  1.0 H 0  100  2 0 03 0 04 0 05 0 06 0 0  Figure 5.3  NORMAL STRESS ON P L A N E OF SHEAR, k P a  Roughness Coefficient vs. Normal Stress Ottawa Sand, Target Sand - Smooth Steel Interface  A  7 08 0 00  65  Figure 5.4a  Interface Friction vs. Shear Displacement Ottawa Sand - Rough Steel Interface (0 - 25 mm)  66  40-i  35-  _  51  0  1  I  l  I  "  I  I  I  I  I  I  I  l  _  |  100 kPa 500 kPa - 750 kPa  1  1  1  1  1  1  1  1  1  1  1  1  50 100 150 200 250 300 350 400 450 500 550 600 650 SHEAR DISPLACEMENT ( m m )  Figure 5.4b  Interface Friction vs. Shear Displacement Ottawa Sand - Rough Steel Interface (0 - 650 mm)  reaches a hmiting value of about 2 6 ° by 100 mm shear displacement while at the low stress 100 kPa 8 has remained essentially unchanged from the peak value of about 22°. For the Target sand-rough steel interface, 8 values illustrated in Figures 5.5a and 5.5b equal approximately 270 at shear displacements of about 5 mm and reach a limiting value of about 29o to 30o by 100 mm shear displacement at high stresses and about 250 at low stresses. The R a values for the rough steel after 650 mm displacement with both Ottawa and Target sands are presented in Figure 5.6.  For both sands R a decreases with  increasing stress indicating that the asperities on the steel surface are being worn. The net result for all stresses except the low stress Target sand interface is a smoother surface after testing.  5.3 Comparison of Surface Performance The coefficient of friction at small displacements JA p , is shown in Figure 5.7 to be dependent on both the surface roughness and the angularity of the interfacing sand. For both sands rough steel mobilizes greater interfacial friction than does smooth steel and the angular Target sand results in a higher value of p:p than the rounded Ottawa sand. The frictional response at very large displacements along the sand-steel interface was shown to be very complex. In order to keep the comparison of the effects of surface roughness and the sand angularity simple, Figure 5.8 has been prepared showing the values of | i as a function of Ra based on high stress tests in which a limiting value of |0. was observed. From this figure it can be seen that at large displacements, the initial roughness of the steel does not affect the value of |i and that the initial angularity of the  40  Figure 5.5a  Interface Friction vs. Shear Displacement Target Sand - Rough Steel Interface (0 - 25 mm)  69  cv  /\ ^ - \ A \ v \  V  V  100 kPa 500 kPa 750 kPa  0  i—i—i—i—i—"—T  i  i  i  i  i  " i  i  i  i  i—i—r  50 100 150 200 250 300 350 400 450 500 550 600 650 SHEAR DISPLACEMENT ( m m )  Figure 5.5b  Interface Friction vs. Shear Displacement Target Sand - Rough Steel Interface (0 - 650 mm)  70  10.0  -  9.0  OTTAWA SAND TARGET SAND  8.0 7.0 6.0 5.0  -  4.0 3.0  ^_R  a  before shearing ~~~  2.0 1.0 0.0 0  -  1 0 02 0 03 0 04 0 05 0 06 0 0  NORMAL STRESS ON P L A N E OF SHEAR, k P a  Figure 5.6  Roughness Coefficient vs. Normal Stress Ottawa Sand, Target Sand - Rough Steel Interface  7 0 8 00 0  71  1.0 0.9 H  £0.8  | 0.7 o  W A R G E T SAND  t—I  OH  ^ 0.6  HOTTAWA SAND  (in  o  A  ^ 0.5 H  A-  2; W  0  0.4  fa  1 o  0.3 H * * * * * OTTAWA SAND A TARGET SAND  0.2 0.1 0.0 0.0  "i  i  i  i  |  0.5  Figure 5.7  i  i  i  i  |  1.0  i  i  i  i  |  1.5  i  i  i—i—|—i—i—i—i—|—i—i—i—r  2.0  INITIAL S U R F A C E ROUGHNESS, R a  2.5  Peak Interface Friction vs. Initial Surface Roughness Ottawa Sand, Target Sand - Steel Interfaces  3.0  72  0.8  H  0.7  , UTARGET SAND  0.6  A -  *—  0.5  HOTTAWA SAND  0.4 H * * * * * OTTAWA SAND TARGET SAND  AA44A  0.0  T—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—r  0 . 5  Figure 5.8  1 . 0  1 . 5  2 . 0  INITIAL SURFACE ROUGHNESS, R a  2 . 5  Residual Interface Friction vs. Initial Surface Roughness Ottawa Sand, Target Sand - Steel Interfaces  3 . 0  sand grains is less significant than at the peak condition. It should also be noted that for both sands the value of u at large displacements is greater than at small displacements. The above findings have significance in field applications such as the insertion of a steel pile or a cone penetrometer into a sand deposit. Meyerhof (1950) presents data from a series of field studies pertaining to the shaft friction mobilized along the length of piles driven in sand. The coefficient of friction K s (Ks=u), was shown to vary from 0.5 to 1.0.  Although the mineralogy and shearing strength of the sand deposits is not  accurately known, these findings are not significandy different from the data presented in this thesis. Current research at the University of British Columbia utilizing a lateral stress sensor with a cone penetrometer allows the direct measurement of 8 during the insertion of the cone. Preliminary results from these data indicate higher 8 values than those observed in this study; however, further analyses are needed to verify these initial findings.  6. CONCRETE INTERFACE TEST RESULTS Tests were performed on concrete specimens prepared using the procedure outlined in section 3.4.2. The surface texture of each specimen varied slightly; however, they could be broadly described as being smooth with occasional irregularities. Figure 6.1 illustrates the friction angle mobilized at the interface of concrete and Ottawa sand at normal stresses of 100 and 500 kPa. The peak friction angle 5p is approximately 27o to 280 and is mobilized by 2 mm relative displacement. This interface friction angle is only lo to 20 below § c v of Ottawa sand. The residual friction angle 5r is essentially coincident with the peak value. Virtually no change in 8 was observed over very large shear displacements up to 650 mm. The friction angle mobilized at the Target sand-concrete interface is presented in Figure 6.2. These curves also indicate that 8 p is very close to the § C \ of Target sand. The slightly erratic data at 100 kPa stress in shear displacement from 0 to 10 mm may be due to the irregular surface of that concrete specimen. The very high 8 values for the sand-concrete interface may be explained with the aid of the photomicrograph presented in figure 6.3.  The photomicrograph clearly  illustrates that after shearing, mortar on the surface of the concrete has been worn away resulting in partially exposed Ottawa sand grains. The consequence of this situation is a sand-sand interface condition which should produce the measured 8 value very close to the <j)Cv of the sand alone. The above findings show that shearing with an angular sand results in a higher interface friction angle than with rounded sand. Concrete pile design should account for  75  40 35 H CO  w  -30 §25 i—*  8o 2 0  £ 15 K W  500 kPa 100 kPa  ~ 10 0  i  0  1  r  "1  I  I  I  I  10  I  I  I  I  I  I—I  1 5  1  SHEAR DISPLACEMENT ( m m )  Figure 6.1  1  1  1  2 0  Interface Friction vs. Shear Displacement Ottawa Sand - Concrete Interface (0 - 25 mm)  1—I  1—  2 5  76  Figure 6.2  Interface Friction vs. Shear Displacement Target Sand - Concrete Interface (0 - 25 mm)  77  the angularity of the interfacing sand and therefore have the computed support capacity and resistance during driving reflect the appropriate 5 values. The data from this research program clearly indicates that for the configurations studied a value of u equal to two-thirds of the shearing strength of the interfacing sand is extremely conservative.  In addition the measured interface friction may also be  dependent on the type of concrete aggregate used although not enough tests were carried out in this program to accurately define this affect.  7. CONCLUSIONS The ring shear apparatus is a useful tool for assessing the friction mobilized at the interface between sand and a variety of construction materials.  Although a shear  displacement gradient in the radial direction exists in this type of apparatus its configuration is far superior to the conventionally used direct shear technique. A better definition of peak and a more confident definition of residual interface friction can be obtained using the ring shear apparatus. The friction mobilized at the sand-geomembrane interface is dependent on the type of geomembrane, its surface texture, the angularity of the interfacing sand and to a lesser extent the normal stress on the plane of shear. The hard, smooth HDPE material exhibits distinct peak and residual 8 values that range from 65% to 55% of <f>Cv for rounded Ottawa sand and 90% to 65% of (j)cv for the angular Target sand respectively. The peak value of 8 is attributed directly to the resistance of the smooth HDPE surface to resist scouring. Roughened HDPE and the much softer smooth PVC geomembrane both show 8 values equal to the § C \ of the interfacing sand. This occurs as a result of the zone of relative shear displacement forming within the sand grains immediately adjacent to the surface  rather than along the sand-geomembrane interface.  The use of a surface  profiling gauge helps explain the relationship between the amount of relative shear displacement and the value of the interface shearing friction for those cases where scouring of the surface occurs. The friction mobilized at small relative displacements between sand and steel is dependent on the surface roughness of the steel and the angularity of the interfacing sand but independent of normal stress. Peak 8 values range from 55% to 70% of (]>cv for the rounded Ottawa sand-smooth steel and angular Target sand-smooth steel interfaces respectively, while for rough steel surfaces 8 values range from 75% to 80% for the same  80 conditions. At very large relative shear displacements, the normal stress appears to have a significant impact on the measured 8 value. A limiting upper bound value of 8 is reached sooner in tests conducted at high stresses than those at low stresses, and it appears that all results approach a common 8 value of approximately 90% of the <t>Cv of the interfacing sand regardless of the initial surface roughness. These findings imply that for piles driven in sand where large relative shear displacements occur at the sand-steel interface no increase in capacity would be realized for an initially rough pile. However, for applications where only small relative displacements at the sand-steel interface are experienced, a rough steel surface will result in a higher 8 value. Tests performed on sand-concrete interfaces indicate that 97% of the <J>CV of the interfacing soil is mobilized at the peak condition independent of the normal stress. In addition 8 values at large relative shear displacements are almost equal to the peak value as a result of the exposed aggregate remaining in immediate contact with the interfacing sand. These data indicate that the commonly used value of 8 equal to two thirds of (|> is very conservative.  REFERENCES Bishop, A.W., Green, G.E., Garga, V.K., Andressen, A. and Brown, J.D. (1971). A New Ring Shear Apparatus and its Application to the Measurement of Residual Strength, Geotechnique 21, No. 4, pp. 273-328. Bosdet, B.W. (1980). The UBC Ring Shear Device, M.A.Sc. Thesis, University of British Columbia. Bowden, F.P. and Tabor, D. (1964). The Friction and Lubrication of Solids, Pt H, Oxford University Press. Brummund, W.F. and Leonards, G.A. (1973). Experimental Study of Static and Dynamic Friction Between Sand and Typical Construction Materials, Journal of Testing and Evaluation, JTEVA, Vol. 1, No. 2, pp. 162-165. Butterfield, R. and Andrawes, K.Z. (1972). On the Angles of Friction Between Sand and Plane Surfaces, Journal of Terramechanics, Vol. 8, No. 4, pp. 15-23. Eigenbrod, K.D. and Locker, J.G. (1897). Determination of Friction Values for the Design of Side Slopes Lined or Protected With Geosynthetics, Canadian Geotechnical Journal, Vol. 24, pp. 509-519. Hailing, J. (1975). Principles of Tribology, Macmillan Education Publ., 1st Ed. Kishida, H. and Uesugi, M . (1987). Tests of the Interface Between Sand and Steel in the Simple Shear Apparatus, Geotechnique 37, No. 1, pp. 45-52. Martin, J.P., Koerner, R.M. and Whitty, J.E. (1984). Experimental Friction Evaluation of Slippage Between Geomembranes, Geotextiles and Soils, Proceedings of the International Conference on Geomembranes, Denver, USA, pp. 191-196. Martin, J.P. and Koerner, R.M. (1985). Geotechnical Design Considerations for Geomembrane Lined Slopes: Slope Stability, Geotextiles and Geomembranes, No. 2, pp. 299-321. Meyerhof, G.G. (1950). The Ultimate Bearing Capacity of Foundations, Geotechnique 2, pp. 301-332. Negussey, D., Wijewickreme, W.K.D. and Vaid, Y.P. (1988). On Geomembrane Interface Friction, Soil Mechanics Series No. 119, Dept. of Civil Engineering, University of British Columbia. Potyondy, J.G. (1961). Skin Friction Between Various Soils and Construction Materials, Geotechnique, Vol. 11 (4), pp. 339-353. Schaffer, G.H. (1988). The Many Faces of Surface Texture, American Machinist & Automated Manufacturing, June, Special Report 801. Uesugi, M . and Kishida, H. (1986). Influential Factors Of Friction Between Steel and Dry Sands, Soils And Foundations, Vol. 26, No. 2, pp. 33-46.  Wijewickreme, W.K.D. (1986). Constant Volume Friction Angle of Granular Materials, M. A.Sc. Thesis, University of British Columbia. Yoshimi, Y. and Kishida, T. (1981). A Ring Torsion Apparatus For Evaluating Friction Between Soil and Metal Surfaces, Geotechnical Testing Journal, GTJODJ, Vol. 4, No. 4, pp. 145-152.  

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.831.1-0062826/manifest

Comment

Related Items