<%BANNER%>

Finite element analysis of a laboratory soil box test facility for evaluating the structural response of concrete pipe

University of Florida Institutional Repository

PAGE 1

FINITE ELEMENT ANALYSIS OF A LABORATORY SOIL BOX TEST FACILITY FOR EVALUATING THE STRUCTURAL RESPONSE OF CONCRETE PIPE By MELISSA KAY CROSBY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Melissa Kay Crosby

PAGE 3

This thesis is dedicated to my loving family, my parents, Sandra and Oler Crosby, my sister Stacy Crosby and to my loving fianc De vin Drake. I would also like to dedicate this thesis to my loving Aunt Linda and Uncle Dewayne as they have offered their support and love throughout this endeavor. It is with the love and support of my family and friends that I am able to reach my goals.

PAGE 4

ACKNOWLEDGMENTS I would like to thank all of the members of my supervisory committee for their help and ideas throughout this effort. Dr. David Bloomquist, committee cochair, provided much insight, knowledge and financial support toward the completion of the work. Dr. Andrew Boyd, committee chair, provided valuable time and knowledge of the subject, as well as financial support, making this research successful. I would also like to thank Dr. H.R. Hamilton for his contribution of time and knowledge, which provided to be invaluable assistance during this effort. An additional thank you goes to Scott Jacobs for his time and the computer expertise he provided during the research. Above all, I would like to thank God for giving me the ability to withstand the trials and tribulations throughout this effort. It is through Him that I am able to persevere and succeed in all my endeavors. I would also like to thank my best friend, who is also my fianc, Devin Drake, for the support and patience he offered me during the research and writing of this thesis. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT......................................................................................................................xii CHAPTER 1 INTRODUCTION........................................................................................................1 2 LITERATURE REVIEW.............................................................................................3 Introduction to Pipes.....................................................................................................3 Concrete Pipes..............................................................................................................3 Concrete Pipe Testing...................................................................................................4 Three-Edge Bearing Strength.......................................................................................6 Bedding Factors and Classifications.............................................................................7 History of Pipe Testing Facilities.................................................................................9 3 STANDARD REINFORCED CONCRETE PIPE VS. FIBER REINFORCED CONCRETE PIPE......................................................................................................22 Background Information on Concrete Pipes...............................................................22 Cracking in Concrete--The Fracture Zone..................................................................23 Standard Reinforced Concrete Pipe--SRCP...............................................................25 Mechanics of SRCP.............................................................................................25 Reinforcement.....................................................................................................27 Strength................................................................................................................27 Structural Performance........................................................................................28 Fiber Reinforced Concrete Pipe--FRCP.....................................................................29 Mechanics of FRCP.............................................................................................30 Fiber-Matrix Bond...............................................................................................31 Fiber-Fiber Interaction.........................................................................................31 Load vs. Deflection in Fiber Reinforced Concrete..............................................32 Stages of Cracking-Fiber Intervention................................................................33 Strength................................................................................................................35 Toughness............................................................................................................35 v

PAGE 6

FRCP versus SRCP.....................................................................................................36 Manufacturing.....................................................................................................36 Installation...........................................................................................................36 Performance.........................................................................................................37 4 INSTRUMENTATION AND TESTING TECHNIQUES.........................................38 Crack Detection and Deflection in Concrete Pipes....................................................38 Background on Acoustic Emission Testing................................................................38 Preamplifiers........................................................................................................42 Postamplifiers and Signal Processors..................................................................42 Transient Recorders.............................................................................................43 Spectrum Analyzers.............................................................................................44 LAM--Local Area Monitor.........................................................................................45 Minolta 3D Digitizer--VIVID 900..............................................................................48 Hardware.............................................................................................................48 Accessories..........................................................................................................49 Operation.............................................................................................................51 5 PLAXIS VERSION 7.2--FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSIS.................................................................................................................53 Introduction.................................................................................................................53 Plaxis--Input........................................................................................................56 Plaxis--Calculations.............................................................................................64 Plaxis--Output......................................................................................................68 6 FINITE ELEMENT ANALYSIS OF A SOIL BOX TEST FACILITY....................74 Introduction.................................................................................................................74 Plaxis 3D--Verification Analysis................................................................................77 Soil Box Analysis.......................................................................................................83 Soil Box Analysis--No pipe.................................................................................84 Soil Box Analysis--Modeled with Test Pipe.......................................................92 Four Wall Friction Analyses...............................................................................99 7 RECOMMENDATIONS..........................................................................................102 APPENDIX PLAXIS 2D ANALYSIS AND RESULTS...............................................105 Dimension Analysis...........................................................................................105 Wall Friction/ Soil Compaction Analysis..........................................................108 LIST OF REFERENCES.................................................................................................112 BIOGRAPHICAL SKETCH...........................................................................................114 vi

PAGE 7

LIST OF TABLES Table page 3.1 Typical Fiber-Matrix Pullout Strengths (Mindess, 2003)........................................32 4.1 Specifications of Minolta VIVID 900 3D digitizer..................................................52 6.1 Material Properties of the soil (Loose & Dense)......................................................76 6.2 Material Properties of the 18 diameter concrete pipes (FRCP & SRCP)...............76 6.3 Material Properties of the 24 diameter concrete pipes (FRCP & SRCP)...............76 6.4 Material Properties of the 48 diameter concrete pipes (FRCP & SRCP)...............77 6.5 Three Dimensional Analysis Verification of Plaxis 2D Wall Friction Analysis.....83 6.7 Displacement of Pipe Length with Shear Stress Induced on the Ends...................101 6.8 Extreme Effective Normal Stresses Along the Length of Pipe with Shear Stress Induced on the Ends...............................................................................................101 A.1 Wall Friction Analysis-Plaxis 2D Finite Element Analysis...................................111 vii

PAGE 8

LIST OF FIGURES Figure page 2.1 Early Concrete Pipe Testing.......................................................................................5 2.2 Three edge bearing test for concrete pipe..................................................................7 2.3 Ohio University Full Scale Testing Site...................................................................13 2.4 The Center for Pipes and Underground Structures Test Facility at Ohio University.13 2.5 Hardie Pipes Rigid Soil Box Front View................................................................15 2.6 Hardie Pipes Rigid Soil Box Full Image View.......................................................16 2.7 Hardie Pipe Flexible Soil.........................................................................................16 2.8 Hardie Pipe Flexible Soil Box Testing conducted at UCF.......................................17 2.9 LVDTs Shown to Measure Deflection of Sidewalls...............................................18 2.10 Flexure Crack At The Crown of Fiber Reinforced Concrete Pipe...........................19 2.11 Reinforced Concrete Pipe Cracking At The Crown.................................................20 3.1 Coordinate system and stress components ahead of crack tip..................................24 3.2 Three Modes of Cracking.........................................................................................24 3.3 Standard Reinforced Concrete Pipe Section Cracked..............................................26 3.4 Standard Reinforcing Steel Rebar............................................................................28 3.5 Typical Fiber Reinforced Concrete Pipe..................................................................29 3.6 Typical load-deflection curve for fiber reinforced concrete in flexure....................33 3.7 Schematic representation of fibers bridging across a crack.....................................34 4.1 Burst Acoustic emission signal with properties.......................................................39 4.2 Acoustic Emission Process.......................................................................................40 viii

PAGE 9

4.3 Pre-amplifiers...........................................................................................................43 4.4 Transient recorder with multiple AE signals............................................................44 4.5 Digital Oscilloscope.................................................................................................45 4.6 LAM--Local Area Monitor......................................................................................46 4.7 Minolta VIVID 900 Non-Contact 3-D Digitizer......................................................49 4.8 Compact Flash Memory Card..................................................................................49 4.9 Rotating Stage Set for Scanning a Full 3-D image..................................................50 4.10 Tripod (left) and Tilting Base Mount (right) for Minolta VIVID 900.....................50 5.1 Plaxis 7.2 Computer Aided Drafting screen used to create modeling analysis........54 5.2 General Settings window in Plaxis 7.2.....................................................................56 5.3 Plaxis 7.2 Main Toolbar...........................................................................................58 5.4 Tunnel Designer in Plaxis 7.2..................................................................................58 5.5 Standard Fixities in Plaxis 7.2 shown on a soil box with right half tunnel..............59 5.6 Material Sets Window in Plaxis 7.2.........................................................................61 5.7 Soil Input in Plaxis 7.2--Mohr Coulomb Model......................................................61 5.8 Beam Properties Input Window in Plaxis 7.2..........................................................62 5.9 Pore Water Pressure & Initial Stress Modes in Plaxis 7.2.......................................63 5.10 Plaxis 7.2 Calculations Program..............................................................................65 5.11 Plaxis 7.2 Calculations Program--Parameters Tab...................................................66 5.12 Plaxis Calculations Program--Multipliers Tab.........................................................67 5.13 Plaxis Output program with deformed mesh displayed on example model.............69 5.14 Plaxis Output Effective Mean Stresses Displayed by Mean Shading......................69 5.15 Plaxis Output Effective Mean Stresses Displayed by Contours..............................70 5.16 Stress Distribution Cross Section A-A in Plaxis 7.2--Output Program...................71 5.17 Horizontal Displacement Cross Section A-A in Plaxis 7.2--Output Program.........72 ix

PAGE 10

5.18 Displacements for the Pipe and Interface--Plaxis 7.2 Output Program...................72 5.19 Bending Moment for the Pipe--Plaxis 7.2 Output Program.....................................73 6.1 Three Dimensional View of Total Displacements for 24 FRCP............................79 6.2 Three Dimensional View of Total Displacements for 24 SRCP............................79 6.3 Left Side Interface of Soil Box Model 24 FRCP...................................................80 6.4 18 Diameter FRCP Friction Area on Surface of Pipe............................................81 6.5 18 Diameter FRCP Near Zero Frictionless Area on Surface of Pipe.....................82 6.6 10 Length Soil Box: Effective Mean Stresses with Sidewall Friction:..................85 6.7 15 Length Soil Box: Effective Mean Stresses with Sidewall Friction....................86 6.8 20 Length Soil Box: Effective Mean Stresses with Sidewall Friction....................86 6.9 10 Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls............87 6.10 15 Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls............87 6.11 20 Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls............88 6.12 10 Length Effective Normal Stresses Left Sidewall Friction Plane.......................89 6.13 15 Length Effective Normal Stresses Left Sidewall Friction Plane.......................89 6.14 20 Length Effective Normal Stresses Left Sidewall Friction Plane.......................90 6.15 10 Length Effective Normal Stresses Left Sidewall Frictionless Plane.................90 6.16 15 Length Effective Normal Stresses Left Sidewall Frictionless Plane.................91 6.17 20 Length Effective Normal Stresses Left Sidewall Frictionless Plane.................91 6.18 10 Length Effective Mean Stress with Friction on Sidewalls.................................93 6.19 15 Length Effective Mean Stress with Friction Sidewalls......................................93 6.20 20 Length Effective Mean Stress with Friction Sidewalls......................................94 6.21 Effective Normal Stress Imposed on Perimeter of Test Pipe...................................94 6.22 10 Length Effective Mean Stress with Frictionless Sidewalls/ Pipe......................96 6.23 15 Length Effective Mean Stress with Frictionless Sidewalls/ Pipe......................96 x

PAGE 11

6.24 20 Length Effective Mean Stress with Frictionless Sidewalls/ Pipe......................97 6.25 10 length Effective Mean Stress with Frictionless Sidewalls and Friction Around Perimeter of Pipe......................................................................................................97 6.26 15 length Effective Mean Stress with Frictionless Sidewalls and Friction Around Perimeter of Pipe......................................................................................................98 6.27 20 length Effective Mean Stress with Frictionless Sidewalls and Friction Around Perimeter of Pipe......................................................................................................98 6.28 Plaxis 3D Test Pipe with Shear Stress Induced on Ends of Pipe...........................100 A.1 Plaxis 2D Symmetry Model of 24 diameter FRCP..............................................105 A.2 Three Different Widths Modeled in Plaxis 2D: 10, 15, 20 feet wide.....................106 A.3 Example of Cross Section Used to Examine the Sidewall Stresses (20 Width)...107 A.4 Example FRCP Cross Section of Sidewall Stresses (10 wide box)......................109 A.5 Example FRCP Cross Section of Sidewall Stresses (15 wide box)......................109 A.6 Example FRCP Cross Section of Sidewall Stresses (20 wide box)......................110 xi

PAGE 12

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering FINITE ELEMENT ANALYSIS OF A LABORATORY SOIL BOX TEST FACILITY FOR EVALUATING THE STRUCTURAL RESPONSE OF CONCRETE PIPE By Melissa Kay Crosby May 2003 Chair: Andrew J. Boyd Cochair: David Bloomquist Major Department: Civil and Coastal Engineering Two and three dimensional finite element analyses were used to examine the stresses on a laboratory soil box test facility designed to evaluate the structural response of fiber reinforced and standard reinforced concrete pipes. Boundary conditions were examined in an attempt to minimize the effects on the concrete pipe tested. The state of stress in the soil due to the applied loading was studied, along with its influence on the concrete pipe. An extensive literature search is presented on the history of soil box testing of pipes. Small and large-scale test facilities have been discussed through previous testing in the literature search. A description of the test facilities, both small and large, is presented. Plaxis, a two and three-dimensional finite element analysis program, analyzes the stresses on the sidewalls for the dimension design of a soil box properly scaled to minimize the boundary effects on the concrete pipe specimens. Comparisons were made among three different proposed box lengths. It was found that as the length xii

PAGE 13

increased, the stress concentrations and intensities decreased, thus minimizing the boundary effects. Sidewall friction and its effects on the test pipe were examined. Two and three dimensional finite element analysis was used to assign a friction angle on the sidewalls in order to model the soil structure interaction. An additional analysis was done to examine the possibility of shear stresses induced on the ends of the pipe. Friction was induced on the front and rear wall to create shear stresses on the ends of the pipe. Shear stresses were induced on the ends of the pipe in the three-dimensional analyses but very little difference in displacement occurred. A comparison of soil backfill, between a loose and dense compaction, showed a large difference in overall settlement. Maximum stresses and pipe deflections were approximately double for the loose compaction when compared to the dense backfill. xiii

PAGE 14

CHAPTER 1 INTRODUCTION Structural performance testing of small diameter buried pipes dates back to the 1930s. Large scale testing of buried concrete pipes provides useful information for evaluating the soil structure response expected under field conditions. In the United States, large-scale test facilities exist at Utah State University, the University of Massachusetts at Amherst, and Ohio University. Additional facilities can be found at the University of Western Ontario in Canada and LGA Geotechnical Institute in Germany (Brachman et al., 2001). Two small-scale test facilities have been designed and constructed by Hardie Pipe, Inc. and are pending patents. Each of the testing facilities at these institutions suffers from limitations stemming from boundary conditions inherent in their respective equipment configurations. Many of the existing test facilities cannot closely approximate expected field conditions with respect to the stress states associated with small diameter buried pipes. When large scale testing is not possible at a particular location, a laboratory test facility is needed to examine the structural response of buried pipes. A soil box test facility is required to investigate small diameter buried pipes and allow a laboratory assessment of their performance under simulated field conditions. Laboratory test facilities allow better control over monitoring and testing procedures than do field testing techniques. Boundary conditions, such as the method of loading and the geometry of the test facility, may significantly influence the results of the test. The structural response of buried pipe can be significantly affected by the boundary conditions if the test facility is not properly designed. Sufficient dimensions are 1

PAGE 15

2 necessary, under expected loading conditions, to limit sidewall deflection and minimize boundary condition effects. The primary concern to be considered in the analysis of testing apparatus design is elimination of the influence of boundary conditions on the structural response of the concrete pipe specimen. Both a two and three-dimensional finite element analysis program (i.e., Plaxis 7.2 and Plaxis 3D Tunnel) will be used to evaluate the boundary conditions induced by an applied distributed load. Stresses imparted by the load dissipate throughout the soil, applying load to the pipe, which is surrounded by backfill. A wall stress and sidewall friction analysis will provide dimensional information that can be used to minimize the effect of boundary conditions. Two different backfills will be evaluated, loose and densely compacted soil. The objective of this work is to perform a finite element analysis for a laboratory test facility to be used in the evaluation of structural performance in small diameter concrete pipes, including both fiber reinforced and traditional steel reinforced concrete. The facility will be fabricated of steel and designed to limit sidewall deflections. The soil box test facility will be loaded using inflatable bladders placed on top of the soil. Attention will be focused on the influence of the boundary conditions on the soil box and how efficiently and accurately the test setup reproduces actual field conditions for small diameter buried pipe. Consideration will be given to determination of optimal test cell dimensions and, the influence of sidewall friction and boundary stiffness on the structural performance of the concrete pipe and soil behavior.

PAGE 16

CHAPTER 2 LITERATURE REVIEW Introduction to Pipes Buried pipes and/or conduits have improved the standard of living for people since the beginning of civilization. Remnants of such structures from ancient civilizations have been found in Europe, Asia, and even the western hemisphere, where some of the ancient inhabitants of South and Central America established functional water and sewer systems (Moser, 2001). Buried pipes serve many purposes, including sewer lines, drain lines, water mains, gas lines, telephone and electrical conduits, culverts, oil lines, coal slurry lines, subway tunnels, and heat distribution lines. In comparing the design used in the 1800s to the design applications we have today, it is apparent that the degree of technology has increased significantly. Engineers and planners take subsurface infrastructure into account before developing buildings and houses for a community. The underground water systems serve as arteries for cities, and the sewer systems serve as veins to carry off the waste (Moser, 2001). High quality drinking water is taken for granted by humans in todays society. To ensure adequate quality, pipes must be designed and constructed to prevent the introduction of contaminants. The same standards apply to sewer pipes so as to prevent seepage of contaminants into the ground, which may reach the water table and aquifers. Concrete Pipes Pipes are classified as either rigid or flexible. A flexible pipe is defined as one that will deflect at least 2 percent without structural distress. Flexible pipes, such as those 3

PAGE 17

4 made of polyethylene plastics, do not fall within the scope of this project. A rigid pipe is one that does not meet the flexible pipe deflection criterion. The two major types of rigid concrete pipes are steel reinforced concrete pipes and fiber reinforced concrete pipes. Parameters of the pipes are analyzed to design for maximum performance. Rigid pipes must have the strength to resist wall stresses due to internal pressures or external loads that are considered critical under anticipated service conditions. Design parameters include strength, stiffness, corrosion resistance, density, durability and ease of joining. A pipe must have sufficient strength and/or stiffness to perform its intended function and also must be sufficiently durable to perform this function throughout its intended service life. Strength is defined as the ability to resist applied stress. Internal pressures, soil pressures, live loads, differential settlement and longitudinal bending moments impose stress on the pipe. Stiffness is described as a materials ability to resist deflection. The modulus of elasticity of a material is directly related to its stiffness, thus affecting deformation of the pipe wall during loading. Durability is the ability of a material to resist degradation (i.e., corrosion, chemical deterioration, abrasion) due to deleterious environmental exposures. Durability is a critical parameter when determining a design service life for performance. Concrete Pipe Testing In the early 1900s, concrete pipe testing consisted of placing sand bags on top of a pipe to obtain a static distributed load as shown in Figure 2.1. In 1913, Marston researched loads induced on buried pipe by the soil above the pipe. This research marked the beginning of the development of a method for calculating earth loads on buried pipes. Shortcomings include an assumption that the vertical load due to the backfill being

PAGE 18

5 Figure 2.1: Early Concrete Pipe Testing (Photo courtesy of Hardie Pipe, Inc.) uniformly distributed. Results were significantly limited by the technology available at the time and did not include such effects as pipesoil interaction or settlement ratio. The Marston load theory, based upon the concept of a prism-shaped soil load being applied to the pipe, resulted in what is known today as the Marston load equation (2.1). 2dddBCW (2.1) W d = load on conduits per unit length C d = load coefficient for ditch conduits = unit weight of backfill B d = horizontal width of ditch at top of conduit As technology progressed, more test methods were standardized for the evaluation of concrete pipe strength. Spangler conducted research in the 1930s that proposed four classifications of bedding for pipes covering normal installations in the field. In order to investigate the soil-structure interaction, the American Concrete Pipe Association (ACPA) undertook a long-range research program to examine the nature of the loading imposed on a buried pipe (Moser, 2001). Their research covered the development of a

PAGE 19

6 finite element program used to simulate non-linear behavior of buried pipe, validation of the program, and analysis of the soil around the pipe. Full scale testing conducted at the Transportation Research Center in East Liberty, Ohio found the strains along the length of the pipe to be insignificant. The results were symmetric about the vertical plane of symmetry of the pipe, thus validating the use of plane strain finite element analysis. Three-Edge Bearing Strength Rigid nonpressure pipes are tested for strength in the laboratory using the three-edge bearing test (ASTM C 497). The performance criteria require the each pipe size and class pipe to reach specified laboratory strengths relative to the anticipated service load condition and required ultimate strength. Traditional design practice uses the three-edge bearing load that produces a 0.01-inch crack width as the design load. The failure load in three-edge bearing test is defined as the load per unit length required to cause crushing or critical cracking of the pipe test specimen. The strength thus obtained is the failure load in the laboratory only, and is not necessarily equivalent to the load that will cause failure in the field under buried conditions. Figure 2.2 shows a schematic diagram of the three edge bearing test for a rigid pipe, where W represents the distributed load, D the pipe diameter, R the radius of the wood support blocks and C the clearance beneath the pipe. Testing of nonreinforced concrete pipes are specified in ASTM C 14. Nonpressure reinforced concrete pipe is specified by its D-load strength, as determined in ASTM C 76. The Dload is defined as the load applied to a pipe under three-edge bearing conditions, expressed in pounds per linear foot per foot of inside diameter.

PAGE 20

7 Figure 2.2: Three edge bearing test for concrete pipe Bedding Factors and Classifications Laboratory testing and field-testing can result in two different strengths for concrete pipes. As previously stated, the strength causing failure in the laboratory does not always cause failure in a buried condition. Past experiments show that the estimated load required to cause failure of a buried pipe, according to the Marston load equation, is greater than that resulting from the three-edge bearing strength. The most important factor influencing this discrepancy is the method in which the pipe is bedded. Bedding factors are variables that were developed to account for the type of soil in which the pipe is installed. The bedding factor (sometimes called the load factor) is the ratio between the strength of a buried pipe and the strength of the same pipe as determined by the three-edge bearing test. Bedding conditions affect the support reaction beneath the pipe and the lateral pressure on the pipe.

PAGE 21

8 Conduits used in ditch drainage have four bedding classifications. The load factors associated with these classifications have been determined empirically and do not take into account any lateral pressures exerted by the backfill. Furthermore, it has been noted that the specified soil compaction cannot be depended upon reliably. Bedding classifications in which a licensed engineer should inspect the installation are Class A, Class B, Class C and Class D. Class A (also known as Concrete Cradle Bedding has a load factor of 2-4 and occurs when the lower part of the conduit is bedded in a cradle constructed of 2,000 psi concrete, having a minimum thickness of one-fourth the pipes internal diameter. The cradle must extend up the sides of the pipe for a height equal to one-fourth its outside diameter. Class B (also known as First Class Bedding) has a load factor of 1.9 and is where the pipe is carefully bedded in an earth foundation, composed of fine granular materials, that is carefully shaped to fit the lower part of the pipe. The minimum bedding width is 60 percent of the pipe diameter and the remainder of the conduit must be entirely surrounded to a height at least 1 ft above its top by granular materials that are carefully placed in order to completely fill all spaces under and adjacent to the pipe. This fill material must be thoroughly compacted on each side, and beneath the pipe, and placed in layers not exceeding 0.5 ft in thickness. Class C (also known as Ordinary Bedding) has a load factor of 1.5, and applies to pipe bedded with ordinary care in an earth foundation shaped to fit the lower part of the pipe, with reasonable accuracy, for a width of at least 50 percent of its outside diameter. The remainder of the pipe is surrounded to a height of at least 0.5 ft above its top by

PAGE 22

9 granular materials that are shovel-placed and shovel-tamped to completely fill all spaces under and adjacent to the pipe (Moser, 2001). Class D (also known as Impermissible Bedding) has a load factor of 1.1. This class is where there is little or no care applied to shaping the foundation to fit the lower part of the pipe or to refill all spaces under and around it. This class of bedding is not recommended for culvert and sewage pipe. Major pipe manufacturing associations recommend bedding factors that correspond to those listed in the Water Pollution Control Federation Manual of Practice, No. FD-5, Gravity Sanitary Sewer Design and Construction (Moser, 2001). History of Pipe Testing Facilities A pipes insitu performance is a function of its material properties, any applied loads, and the soil-structure interaction. Concrete pipe testing dates back to the early 1900s, when testing involved the placement of sand bags on top of the pipe to achieve static, distributed load. Traditional concrete design methods are based on research conducted at Iowa State University that dates back to the 1930s. By the 1930s, Spangler had proposed four classifications of bedding that covered the range of installations that could be anticipated in normal installations (Parmalee, 1977). Some design methods were found to be too conservative as research progressed. Prior to the 1970s, the testing of buried pipes, along with the soil-structure interaction, did not produce accurate results. In 1970, the American Concrete Pipe Association undertook a long-range research program to determine the nature of loading imposed on buried concrete pipe and to develop a reliable design method based upon soil-structure interaction (Parmalee, 1977). This research program purported to develop a comprehensive finite element analysis program to simulate the non-linear behavior of

PAGE 23

10 buried concrete pipe. The results of the research program indicated that strains along the length of the pipe were insignificant. It was also determined that results were symmetric about the vertical plane of symmetry of the pipe, thus validating the use of plane strain finite element analysis. Soil box testing began in the 1960s. This approach consists of a three-dimensional box of known size, containing a pipe and backfilled with soil. Between 1960 and 2000, several researchers conducted large and small scale testing on pipes under simulated insitu conditions. Each of these research programs were formulated so as to design a test that would yield results similar to those found in the field. Selander et al. evaluated reinforced plastic mortar pipe in 1972. His design for the height of overlying soil (approximately 12 feet high) proved later to be overly conservative. Bland and Sheppard (Bland, 1985) used research findings from the Transport and Road Research Laboratory and the Clay Pipe Development Association obtained with unrealistic boundary conditions to develop a test for clay pipes using a large test pit in order to investigate structural performance. By using a large test pit, the prior limitations of unrealistic boundary conditions found in a small soil box test were eliminated. Bishop tested buried PVC pipes in a soil cell and in a full-scale embankment. He proved it is possible to separate the long-term behavior of the pipe from the long-term behavior of the soil. In-situ load increases with time, whereas the loads applied to the soil cell will decay since the PVC pipe undergoes stress relaxation (Bishop, 1981). In 1981, Gaube and Miller designed a sand box with 5mm thick sheet steel to test plastic sewer pipe. They placed a water-filled bladder between the tops of the soil and

PAGE 24

11 the soil box lid to apply pressure to the top of the soil. Their design employed a box width of eight times the diameter of the pipe. The soil box tests produced results similar to field tests and indicated that plastic pipe could withstand up to 150 feet of earth fill, assuming that compacted soil was used for the fill (Gaube, 1981). Molin tested flexible 200mm diameter PVC pipes in a soil box, while investigating backfills of sand, compacted clay and uncompacted clay (Molin, 1981). His results showed that the vertical soil pressure above the pipe increases with increasing pipe stiffness. Molin stated that his soil box test results were similar to field tests yet claimed the field tests were more vague. Measured strains were compared to calculated strains and found to be acceptable. In 1985, Singhal and Veliz believed that the boundary conditions and edge effects could be eliminated (Singhal, 1985). In other words, the soil surrounding the pipe in the soil box carried stresses and strains that dissipated laterally with distance. Singhal and Veliz tested cyclic torsion, axial pullout, and bending on buried pipes. In the same year Todres and McClinton used a soil box constructed of 19mm plywood panels pinned through steel channels and angles. A 4-inch diameter steel pipe was tested for performance by measuring strains on the pipes walls. A controlled load placed on top of the fill allowed them to compare the measured bending stresses with calculated stresses, and obtained a reasonable correlation (Todres, 1985). A finite element analysis was implemented into the design of a soil box test facility (Zanzinger, 1995). Zanzinger and Gartung based their design on the drop in modulus of elasticity of the pipe over time. When loading the surface of the soil, stresses from the pipe are distributed to the soil surrounding the pipe. A finite element analysis approach

PAGE 25

12 was used to determine the required soil box width needed to eliminate stresses acting on the sides of the box. During the 1000-hour test, a laser was used to measure pipe deformation. The size of a testing facility determines whether one needs to design a laboratory test site to simulate field conditions. Some research centers are equipped with full-scale facilities that reproduce results that would be expected in the field because the research is essentially performed in a field test site. The Center for Pipes and Underground Structures was developed by both the Ohio University and ORITE (Ohio Research Institute for Transportation and the Environment) and is shown in Figures 3 and 4. This facility is one of the largest existing test facilities of this nature. When a large test site like this is not available, a smaller facility or soil box is needed to recreate tests under laboratory conditions. Laboratory tests provide better control of the test and conditions. As stated earlier, Ohio State University is home to one of the largest in situ pipe testing facilities. Smaller test facilities are needed for researchers unable to gain access to such large test pits. Brachman et al. states that boundary conditions, such as the geometry of testing facilities and the method of load application, may significantly affect test results. Advantages of soil box testing include better control and access for instrumentation.

PAGE 26

13 Figure 2.3: Ohio University Full Scale Testing Site (Courtesy of James Hardie Pipe, Inc.) Figure 2.4: The Center for Pipes and Underground Structures Test Facility at Ohio University. (Courtesy of James Hardie Pipe, Inc.)

PAGE 27

14 Brachman et al. conducted research entitled Interpretation of Buried Pipe Test: Small-Diameter Pipe in Ohio University Facility. A small diameter leachate collection pipe was analyzed using two and three-dimensional analyses. Numerical analysis provides one way to assess boundary conditions on measured results when laboratory tests are conducted. Tests were performed at the Ohio University facility for small diameter high-density polyethylene leachate collection pipes. Boundary conditions of the test facility, along with the stress states in the soil and the pipes response to the soil interaction, were investigated. Two large hydraulic cylinders were used to apply a vertical force to a loading platform, thus loading the soil and underlying pipe. The backfill surrounding the pipe consisted of crushed stone over clay bedding to simulate a leachate collection system. The most important boundary condition for this study was the method of load application. A platform of eight W-shaped steel beams welded together provided a rigid footing for the applied load. Results from the finite element analysis clarified the state of stress in the soil due to the overburden load induced by the platform loading. The facility results were compared to the expected results from the field installation. The tests results were complex and required careful interpretation before drawing any conclusions concerning pipe performance for leachate collection in landfills. The stresses induced by the rigid platform differed from the expected uniform load in a landfill. It was found that, at low load levels, a portion of the backfill material at the soil-pipe interface yielded due to the crushed stone behaving as a beam in bending. This effect caused a reduction in the lateral support provided to the pipe, thus increasing pipe deformations and altering the mode of pipe deflection. The measured deflections were

PAGE 28

15 higher than expected in a landfill situation. Care must be taken when interpreting the results from facility testing. Numerical analysis can be successful when correctly interpreted to evaluate boundary conditions. James Hardie tested a rigid soil box constructed with lateral sidewalls of 9mm angle iron. The angle iron effectively restricted any lateral movement. Hardies rigid pipe testing apparatus is shown in Figures 2.5 and 2.6. After realizing that the rigid box was restrictive and did not accurately represent in situ conditions, Hardie developed a flexible soil box (shown in Figure 2.7). The new box was designed with moving lateral walls. Leaf springs supported the walls in an attempt to simulate in situ stiffness. The flexible soil box provided the option of changing the sidewalls lateral stiffness to reflect different burial conditions. Figure 2.5: Hardie Pipes Rigid Soil Box Front View (Courtesy of Hardie Pipe Inc.)

PAGE 29

16 Figure 2.6: Hardie Pipes Rigid Soil Box Full Image View (Courtesy of Hardie Pipe) Figure 2.7: Hardie Pipe Flexible Soil (Courtesy of Hardie Pipe, Inc.)

PAGE 30

17 Testing using this type of flexible soil (performed by Hardie Pipe) was observed on November 13, 2002 at the University of Central Florida (UCF) in Orlando, Florida as seen in Figure 2.8. This research will compare the soil box test results to in situ results. Two different types of reinforced concrete pipe were tested; Hardie Pipes fiber reinforced concrete pipe (in both dry and saturated conditions) and standard reinforced concrete pipe. The backfill used in these tests was compacted coarse sand. Two sheets of Teflon were used to line the inside walls of the box in order to reduce the sidewall friction angle to less than 5 degrees. The box was loaded using a concrete slab placed on top of a series of parallel 2 x 4s in order to simulate a uniformly distributed load. The loading was controlled and monitored using a personal computer. Figure 2.8: Hardie Pipe Flexible Soil Box Testing conducted at UCF.

PAGE 31

18 During the test, LVDTs were used to measure the deflection of the sidewalls induced by the applied loading (Figure 2.9). The load level was noted at the point where the first crack visible to the naked eye formed and loading was then continued until multiple cracks appeared. The flexible soil box is equipped with a viewing port of plexi-glass to monitor the pipe during loading. Interior surfaces of the crown and invert of the pipe are visible from the viewing port. Cracking of the fiber reinforced concrete pipe is Figure 2.9: LVDTs Shown to Measure Deflection of Sidewalls. shown in Figure 2.10, as viewed through the viewing port of the soil box. Each of the tests was continued until cracks propagated across the crown and invert of the pipe. For both the dry and saturated conditions of fiber reinforced concrete pipes, the pipe was loaded until cracking occurred in flexure.

PAGE 32

19 Figure 2.10: Flexure Crack At The Crown of Fiber Reinforced Concrete Pipe. Standard reinforced concrete pipes were also tested using the flexible soil box. Compaction of the backfill was performed in accordance with the same procedure as used in the fiber reinforced concrete pipes. Cracking of the reinforced concrete pipe occurred at the crown of the pipe, as shown in Figure 2.11. Though the same loading procedure was used for the reinforced concrete pipes, it resulted in a much lower maximum load when compared to the fiber reinforced concrete pipes. The testing observed on November 13, 2002, was only a portion of the testing program being conducted at the University of Central Florida.

PAGE 33

20 Figure 2.11: Reinforced Concrete Pipe Cracking At The Crown. Brachman et al. discussed the design of a laboratory facility and the testing of buried pipe performance. Their study considered a limiting applied pressure of 1000kPa, based on a burial length of 50 m and a soil density of 20kN/m 3 and then used finite element analysis to determine the effect of sidewall friction on the soil. Symmetry about the vertical diameter of the pipe was assumed so that only one half of the test facility need be modeled. The load was applied as a uniform pressure and the soil box contained a 2,000 mm 2 soil prism that extended to a height of 1,600 mm. These dimensions allowed only small horizontal deflections under large vertical pressures. The large distance between the pipe and the sidewalls was an attempt to provide lateral stiffness, while at the same time avoiding alteration of the pipes behavior. Hardie stated that the effect of sidewall friction should be considered with respect to the pipes response and not to the soil box wall. Brachman et al. stated that sidewall

PAGE 34

21 friction reduces the amount of load experienced by the pipe, thus resulting in a reduction in vertical deflection of the pipe. Based on a finite element analysis, Brachman et al. concluded that a sidewall friction angle of 5 degrees best simulates in situ conditions. In order to obtain a friction angle of 5 degrees, Brachman used polyethylene sheets lubricated with DC44 silicone grease. His research highlighted the importance of recognizing that the pipe distributes both horizontal and vertical stresses and that a reasonable model of the soil stresses can be achieved when the top and bottom of the soil is at least a distance of one diameter from the pipe. Research conducted at the University of Florida will provide a bedding depth below the pipe of one pipe diameter. In the early years of soil box testing, the boundary conditions induced by the equipment caused poor correlation with in situ condition results. Boundary conditions were later revised to better represent such conditions. From the literature review conducted, it has become apparent that boundary conditions and pipe installation technique are of high importance when trying to simulate in situ performance with pipes tested in a soil box. Therefore when designing a soil box, controlling boundary conditions and soil stiffness is fundamental to producing accurate and acceptable results.

PAGE 35

CHAPTER 3 STANDARD REINFORCED CONCRETE PIPE VS. FIBER REINFORCED CONCRETE PIPE Background Information on Concrete Pipes Presently, concrete pipes are fabricated in a variety of sizes, ranging from 4 inches to over 16 feet of inner diameter. Concrete pipes are used to transport liquids under gravity flow and are implemented as highway culverts, storm drains and sanitary sewers. The evaluation of concrete pipes for use as storm drains under gravity flow is the focus of this project. Concrete pipes are reinforced against crushing when the inner diameter is greater than 24 inches. Standard reinforced concrete pipes (SRCP) are fabricated in accordance with ASTM C76 Standard Specification for Reinforced Concrete Culvert, Storm Drain, and Sewer Pipe. ASTM C76 covers a size range diameter from 12 inches to 108 inches, with an exception for larger pipe diameters. SRCP are heavy products and require lifters capable of proper placement and installation. The SRCP used for this research will come from Rinker Materials. Hardie Pipe introduced fiber reinforced concrete pipes (FRCP) into the civil construction market for large drainage pipes in early 2002 under the companys trademark Fiber Reinforced Concrete Speed Drain Pipes. The concrete properties for each type of pipe include compressive strength, density, absorption, water-cement ratio, cementious materials, aggregates and versatility. 22

PAGE 36

23 Cracking in Concrete--The Fracture Zone Cracking in concrete begins at the micro level. The fracture zone is defined as the state when the stress or the strain is increased to the point where the atomic bonds within the matrix are broken and the solid material is cracked or fractured. As the stress load increases, cracks will develop and propagate. In concrete, cracks typically propagate due to tensile stresses, as a result of the low tensile strength of concrete. Though tensile strength is not considered in concrete design, is the cause of most crack initiation and propagation. Failure of concrete is due to tensile stresses induced by loads and/ or environmental changes. Concrete failure is often the end result of microcracking associated with the interfacial region between the hydrated cement paste and aggregates or other inclusions (reinforcing steel, fibers, etc.). Cracks are initially localized but increase in size as the applied stress is increased. In certain circumstances, cracks can propagate very fast. Cracks propagate in three different modes. Cracking modes are classified as either plane strain modes or anti-plane strain modes. Mode I and Mode II deformation are plane strain and Mode III is anti-plane strain. The deformation of the crack is discussed using the coordinate system shown for Mode I displacement in Figure 3.1 below. A description of the three modes of cracking will follow the coordinate system shown in Figure 3.2. In Figure 3.1, an isotropic solid is shown with the origin of the coordinate system at the tip of the crack. It is important to note that the solid shown in Figure 3.1 represents an isotropic solid only. Anisotropic solids are quite complicated when analyzing the fracture of the solid. Mode I deformation occurs when a transverse plane stress is applied to the crack forcing the crack to open up along the y-axis. Mode I is the most prevalent type of cracking in a brittle material. Mode II deformation is a result of a shear stress applied on

PAGE 37

24 the cracked solid that forces the faces of the crack slide over one another parallel to the xy plane. Mode III occurs when a shear stress is applied resulting in the crack faces sliding over each other perpendicular to the xy plane. Cracking of concrete occurs in different stages beginning at the micro level. Figure 3.1 Coordinate system and stress components ahead of crack tip (Mode I displacement) (Mindess, 2003). Figure 3.2 Three Modes of Cracking (Mindess, 2003).

PAGE 38

25 Standard Reinforced Concrete Pipe--SRCP SRCP is widely used in the construction market today. Uses include highway culverts, storm drains and sanitary sewers. Pipes without reinforcement are used where the application is suitable for such products. The majority of applications require reinforcement in order to increase overall strength and resist loads applied over the service life of the pipe. For example, a highway culvert located under a bridge would be subjected to cycling of live loads from traffic. Reinforcement would increase the amount of live load capability prior to failure. SRCP uses steel rebar for reinforcement. The rebar is oriented within the pipe as a longitudinal spiral or linear section placed around the perimeter, parallel to the length of the pipe. Welded wire mesh is also used for additional reinforcement, allowing for increased bonding of the concrete to the rebar reinforcement. An SRCP is shown in Figure 3.3 with the steel reinforcement visible. This pipe section was tested in Hardie Pipes soil box apparatus located at the UCF. The aggregates are also visible, along with the cracking induced by the load applied. Properties that will be discussed for SRCP include durability, such as the ability to resist corrosion of rebar and proper bonding of the concrete to the rebar. Without sufficient bonding of the concrete and rebar, the reinforcement cannot properly carry the tensile stresses and prevent cracking of the concrete matrix. Mechanics of SRCP Different sizes of steel rebar are used to reinforce cement based concrete structures. In this research, concrete pipes will be tested for ultimate strength. Concrete is weak in tension, strong in compression. When tensile stresses are induced in the

PAGE 39

26 Figure 3.3 Standard Reinforced Concrete Pipe Section Cracked. concrete, the use of steel bars or wires carries the tensile stresses. Sufficient bond between the concrete and steel is necessary to allow transfer of the tensile stresses to the steel. The structural performance of SRCP thus relies on the bond between the steel and concrete. Steel design for concrete structure is specified in ACI 318. Steel rebar left unprotected from the environment will corrode, resulting in a loss of strength. In reinforced concrete structures, the steel rebar is placed within the concrete matrix to provide protection from corrosion.

PAGE 40

27 Reinforcement The most commonly used reinforcements for non-prestressed members are hot-rolled deformed bars and wire fabric. Hot-rolled deformed steel bars are basically round in cross section with lugs or deformations rolled into the surface to aid in developing anchorage or bond with the concrete. Steel reinforcing bars are manufactured according to ASTM specifications: ASTM A615-85 Specification for Deformed and Plain Billet-Steel Bars for Concrete Reinforcement; ASTM A616 Specification for Rail-Steel Deformed and Plain Bars for Concrete Reinforcement; ASTM A617 Specification for Axle-Steel Deformed and Plain Bars for Concrete Reinforcement; ASTM A706 Specification for Low-Alloy Steel Deformed Bars for Concrete Reinforcement. Strength Reinforcing steel bars are manufactured in three grades; 40, 50, and 60, with yield strengths of 40, 50 and 60 thousand pounds per square inch (ksi). The 40-ksi bar is the most ductile of the three. Hot rolled steel bars are shown in Figure 3.4. Steel reinforcement is classified by its nominal diameter, expressed in eights of an inch. ASTM specifications for reinforcing bars define the yield strength as the stress carried by the bar at a strain of 0.005. ACI Sections 3.5.3.2 and 3.5.3.4 to 3.5.3.6 define the yield strength as the stress carried at a strain of 0.0035. ACIs definition is based on the strain at which concrete crushes when bars are in compression, a situation where a strain of 0.005 may never be reached. ASTM specifications for yield strength are based on mill tests that are carried out at a high rate of loading.

PAGE 41

28 Figure 3.4 Standard Reinforcing Steel Rebar. Structural Performance Structural performance is affected by several factors, such as bond, temperature and surrounding environmental conditions. The bond between the concrete matrix and the steel defines the stress transfer within the concrete. Tensile stresses are absorbed by the structural steel, providing tensile strength to the concrete structure. Debonding of the concrete and steel results in lower overall strength and performance of the concrete. Temperature affects the strength and performance of steel. The concrete cover over the steel resists heat penetration into the concrete, preventing temperature rise and decreased strength in the steel. Steel temperatures exceeding about 850F result in a significant drop in both yield and ultimate strength. Adverse environmental conditions can cause corrosion of the steel reinforcing bars, thus decreasing performance. Water and oxygen penetration into the concrete from the surrounding soil or water increases the possibility of steel corrosion, especially in the presence of chlorides. Concrete pipe serving as storm drains are susceptible to water penetration from both inside and outside the pipe. Increasing strains applied to the pipe will induce stresses and result in cracks forming throughout the pipe. Once a crack occurs, channels are created throughout the concrete, allowing water to enter at a much

PAGE 42

29 higher rate and accelerating corrosion of the steel reinforcement. Once the steel corrodes, the strength of the concrete structure is impaired. Fiber Reinforced Concrete Pipe--FRCP The fibers in fiber reinforced concrete pipe serves as the reinforcement used to carry tensile stresses. A diameter range of 12 to 48 is produced, with a standard length of 16. Pipe strengths within each size category are divided into five different classes, based upon wall thickness (I V, with I being standard and V extra heavy). FRCP is manufactured in accordance with ASTM Standard C1450 and FDOT Standard 941, with designs from AASHTO Section 17 or LRFD Section 12. A typical FRC pipe is shown in Figure 3.5. Figure 3.5 Typical Fiber Reinforced Concrete Pipe

PAGE 43

30 Mechanics of FRCP Different types of fibers are used to reinforce cement-based matrices. Fibers offer a more economical means of reinforcement for concrete structures of small size. Economically, fibers are cheaper than the usual steel reinforcing bars used, although fibers cannot be considered a replacement for traditional structural reinforcement of massive structures. The ACI 318 design code is based solely on concrete strength as the design criteria, considering only the peak loads a structure can withstand. Fibers are used very little in conjunction with structural steel due to the lack of consideration for post peak behavior. Fibers are most effective in the post peak phase of loading, since the fibers do not carry significant load until after the first crack occurs. They thus do no affect the first-crack strength of concrete and have very little effect on ultimate load. Fibers are generally randomly distributed throughout the cross section. Most fibers are short and closely spaced within the matrix to allow formation of a bond between the fiber and the matrix. Stresses are sometimes transferred from fiber to fiber within the matrix, thus resulting in fiber-fiber interaction. Fibers are primarily used to control crack propagation by bridging across cracks as they begin to open when the concrete strain exceeds its ultimate capacity. An important factor used to determine fracture properties of a material is the stress intensity factor. Cracking occurs after the critical value of the stress intensity factor is reached. Without going into the detail necessary to solve for the stress intensity factor, a simple overview will be provided to explain its importance. The stress intensity factor K is considered to be a single-parameter description of the stress and displacement fields in the region of a crack tip (Mindess, 2003). When the

PAGE 44

31 stress intensity factor reaches a critical value, unstable fracture will occur, resulting in cracks forming. Fibers contribute to a more ductile concrete matrix. Fracture toughness is increased due to the bridging of fibers across the crack as they resist crack opening. As the applied load is increased, fiber reinforcement is designed to absorb the maximum amount of energy possible before unstable behavior of the matrix occurs. Fiber-Matrix Bond In properly designed cement composites for maximum performance, fibers are randomly dispersed throughout the matrix. The fiber matrix bond is a function of both the fiber and matrix properties. Fiber-matrix bond strength is determined from fiber pullout tests and reported as an average value over fiber surface area. Stresses are induced on the fiber-matrix bond, resulting in fiber pullout or debonding from the matrix after the maximum strength is reached. Typical fiber-matrix pullout strengths are shown in Table 3.1. When using cement as a matrix, the fiber-cement interface can become complicated if a chemical reaction occurs between the cement and fiber. Formation of water around the fibers can occur due to bleeding in fresh concrete or insufficient packing of cement grains around the fibers. Under such conditions, the matrix becomes porous near the surface of the fibers than in the bulk cement paste. Fiber-Fiber Interaction A high ratio between fiber modulus of elasticity and matrix modulus of elasticity facilitates stress transfer from the matrix to the fiber. Fiber to fiber interaction occurs when stress is transferred between fibers. Fibers take on the tensile stress induced within the fibrous composite material. Stress concentrations will arise at the fiber ends if they

PAGE 45

32 are discontinuous. The tensile stress that would be assumed by the fiber without the discontinuity must be taken up by the surrounding fibers in the composite. Table 3.1 Typical Fiber-Matrix Pullout Strengths (Mindess, 2003). Matrix Fiber Pullout Strength Mpa (lbs/in2) Cement Paste Asbestos 0.8-3.2 (115-460) Glass 6.4-10.0 (930-1450) Polycrystalline alumina 5.6-13.6 (810-1970) Steel 6.8-8.3 (990-1200) Mortar Steel 5.4 (780) Concrete Steel 3.6 (520)(first crack) 4.2 (610)(failure) Nylon 0.14 (20) Polypropelene 1 (150) The effect of fiber-fiber interaction on stress transfer is described by Rileys theory (Beaudoin, 1990), which states that discontinuous fibers can contribute a maximum of only 6/7 of their strength to the strength of the composite, decreasing to 1/2 for badly flawed fibers. Load vs. Deflection in Fiber Reinforced Concrete As the applied load on a material increases, the strain will increase as well resulting in deformation and deflection. A typical stress-strain curve for fiber reinforced concrete is shown in Figure 3.6. Point A indicates load at which the first crack occurs in the matrix, known as the first-crack strength. The stress at which the first crack occurs is the same in fiber reinforced concrete, traditionally reinforced concrete and plain concrete. The strength of fiber reinforced concrete in the post-cracking zone comes from the transfer of loads across the cracks by the fibers, increasing the strength of fiber reinforced concrete over that of the matrix. Fibers increase the toughness by providing energy absorption mechanisms through the gradual debonding and pull out of the fibers bridging across the cracks (Mindess, 2003).

PAGE 46

33 Figure 3.6 Typical load-deflection curve for fiber reinforced concrete in flexure (Mindess, 2003) Stages of Cracking-Fiber Intervention The first stage of cracking is the development of microcracks. Microcracks form due to high stress concentrations in specific regions of the concrete matrix. The microcracks lengthen, meet and coalesce to form one or more macrocracks. The final stage corresponds to the propagation of cracks. Much research has been done to find out how fibers intervene during the various stages of cracking. During the first stage, the fibers respond to the uniformly distributed microcracking creating a stitching effect on the micro-cracks and preventing propagation. The intervention of the fibers retards the microcracking coalescence phase and the creation of macro-cracks. Once macrocracks develop, the fibers will bridge across them and serve as reinforcement similar to the steel in reinforced concrete. This bridging mechanism is illustrated in Figure 3.7.

PAGE 47

34 Figure 3.7 Schematic representations of fibers bridging across a crack (Mindess, 2003). Fibers bridge across the cracks as they open within the matrix. The stress field around the crack is shown along with the traction-free crack length, fiber bridging length and the aggregate interlock. Stresses are absorbed by the fibers in three different areas shown in figure 3.7. The traction free zone is where fibers have pulled out due to the crack opening up wide enough to overcome the pull out strength. Stresses are absorbed by the fibers and transferred across the fibers by frictional slip in the fiber-bridging zone. Aggregates also absorb energy as their interlocking is distributed to the matrix itself in the microcracked matrix process zone. As stated previously, fibers intervene at two levels; the material level (during macrocracking) and at the structural level (during stress redistribution). In order for the fibers to respond effectively, fiber dimensions and properties must be optimized for the matrix material. The volume proportion of fibers also needs to be optimized to provide maximum mechanical performance. Two approaches are considered during the mix design stage. Since only discontinuous fibers are used, one possibility is to mix in a high percentage of relatively short fibers, resulting in an increase in the strength of the FRC, as the dimensional scale of the fibers is the same as that of the microcracks. Another

PAGE 48

35 possibility is to mix in a low percentage of long fibers, resulting in improved ductility of the structure. This is effective, with regard to macrocracking, since the fibers have sufficient anchorage length on either side of the crack as it opens across the fibers. Strength The role of fibers in concrete is not to increase the overall strength. Some minimal increase in strength will occur in compression and flexure. Research shows that in direct tension, where it would be expected that fibers should be most effective in terms of strength, strength increases are limited to about 30% for a steel fiber volume of 1.5% (Mindess, 2003). Fibers are reported to have no major effect on shear and torsional strength or on elastic modulus. Toughness Fibers have an enormous effect on toughness. If the fibers possess sufficient strength and stiffness, and bonded well with the matrix, they will minimize cracking. This allows the fiber reinforced concrete to withstand significant stresses over a relatively large strain capacity in the post-cracking (or strain-softening) stage, thus providing a considerable amount of post-cracking ductility (Mindess, 2003). Certain fibers have a greater effect on increasing the toughness of fiber reinforced concrete. Deformed fibers tend to have a greater effect on increasing toughness since they bond to the matrix better, increasing the overall pullout strength. Increasing the bond strength past the point where the fibers themselves begin to fail in tension, however, is counterproductive. More energy is required to pull a fiber out of the surrounding matrix than to actually break the fiber across a crack. Thus, designing fibers and FRC for fiber pullout, instead of breakage, is the key to maximizing energy absorption and ductility during failure.

PAGE 49

36 Mindess shows that steel fibers are more effective that polypropylene fibers in improving toughness because of their higher stiffness. FRCP versus SRCP Performance of a pipe is based on its ability to maintain shape under service loads, resist cracking from applied stresses, and resist deterioration of the pipe material due to environmental exposure. Concrete drainpipes are subjected to a number of deterioration mechanisms. Stresses will induce microcracking that eventually coalesces into macrocracks within the concrete matrix. These cracks provide openings for water and other deleterious materials to penetrate into the concrete. In traditional steel reinforced concrete pipe, such ingress can induce corrosion of the steel reinforcement, leading to a decrease in the ultimate strength of the pipe. Though not susceptible to corrosion, the durability concerns related to cellulose fiber reinforced concrete pipe have not yet been fully investigated. A comparison of the properties and characteristics of FRCP and SRCP will be discussed. Manufacturing FRCP is manufactured using high-pressure autoclaving, allowing higher concrete strengths to be obtained for a given set of ingredients. SRCP is manufactured with a low water to cement ratio concrete mix, which is cast into forms during an automated fabrication sequence. Installation The unit weight of a concrete pipe has a significant impact on the installation process. A lighter pipe is easier and quicker to install than a heavy pipe. For a given size and class, an FRC pipe is about half the weight of a corresponding SRC pipe, making it easier to install and handle.

PAGE 50

37 The standard length of SRCP is six feet. FRCP, on the other hand, comes in a standard length of 16 feet. The installation work required for FRCP, compared to SRCP, is cut in half due to this longer standard length. FRCP is also easier to cut when length adjustments are needed, requiring less time than SRCP. An important part of the installation of concrete pipes is the joint seal at the ends of the pipes. Both pipes use rubber gaskets at joints to provide a tight seal, preventing the flow of liquid into, and out of, the pipe system. Performance A vital property affecting mechanical performance is the bond of the concrete to the reinforcement. Steel rebar is more difficult to bond to than cellulose fibers, which are randomly dispersed throughout the concrete mix. This random dispersion allows for increased bonding due to the small size of the fibers and the increase in total surface area inherent in smaller particle sizes. Even with these improvements in bond performance, it still must be remembered that fiber reinforcement cannot be considered a complete replacement for the reinforcement of concrete structures. Both FRCP and SRCP will be tested in the soil box designed by the University of Florida for research funded by the Florida Department of Transportation. The FRCP will come from Hardie Pipe, Inc. while Rinker Materials, Inc will provide the SRCP. Following design and construction of the soil box testing apparatus, each type of pipe will be evaluated as to its insitu performance.

PAGE 51

CHAPTER 4 INSTRUMENTATION AND TESTING TECHNIQUES Crack Detection and Deflection in Concrete Pipes During the testing, the two primary parameters, other than applied load, that will be monitored include cracking and deflection. The ultimate failure of a pipe specimen will eventually be define relative to one or both of these serviceability limits. There are many test methods used today to detect cracks and measure deflection of a structure under load. Determining which of these to implement comes down to applicability, accuracy, and ease of use. Since the pipes are to be tested in a buried condition, detecting cracks occurring on the outer face of the pipe may prove problematic. With current advances in technology, crack detection is more easily accomplished using acoustic emission monitoring techniques. This should allow detection of crack formation without the need to actually observe the crack. The deflection of the concrete pipes is an observation of the pipes deformation while subjected to an applied load. A novel method will also be used to observe and record pipe deflection during testing, specifically three-dimensional imaging of the inner surface of the pipe with a laser based digitizing system. Background on Acoustic Emission Testing Acoustic emission is defined as an acoustic wave generated by a material when subjected to an external stimulus causing an irreversible change in the material. There are two types of acoustic emission signals, continuous and burst signals. A continuous 38

PAGE 52

39 emission is a sustained signal level produced by rapidly occurring emission events such as plastic deformation. A burst emission is a discrete signal related to an individual emission event occurring in the material, such as a crack forming or propagating in a brittle material, such as concrete. An acoustic emission burst signal is shown in Figure 4.1. The term acoustic emission signal is often used interchangeably with simply acoustic emission. An acoustic emission signal is defined as the electrical signal received by the sensor in response to the acoustic wave moving through the material. This emission is picked up by the sensor and transformed into an electrical signal, then analyzed by acoustic emission instrumentation, resulting in information about the material that generated the emission. Figure 4.1: Burst Acoustic emission signal with properties Acoustic emission is a passive, non-destructive monitoring technique. This means that there is no input from an outside source; the technique purely monitors the material being tested. Acoustic emission is typically used to detect cracking, delamination (slip between concrete and steel reinforcement), failure of strands in prestressing tendons, and

PAGE 53

40 fracture or debonding of fibers in fiber reinforced concrete. A typical acoustic emission system setup is shown in Figure 4.2. Figure 4.2: Acoustic Emission Process Equipment and Instrumentation An Acoustic emission system includes at least one sensor and a preamplifier. Most systems also include postamplifiers and signal processors. The system to be used in this research project is a LAM a Local Area Monitor that consists of eight sensors, which will be mounted on the pipe during load testing. More specialized equipment often associated with such a system includes transient recorders, spectrum analyzers, distribution analyzers and spatial discrimination circuits (Beattie, 1993). Microprocessor based systems have become more widely used in recent years that can perform single channel analysis along with source location for up to eight AE channels. Sensors The Acoustic emission sensor is the most important part of the instrumentation and must be properly mounted to assure the required sensitivity. Selection of sensors and or transducers will depend on test parameters and desired results. Sensors are calibrated

PAGE 54

41 using test methods stated by societies. ASTM states that annual calibration and verification of pressure transducer, AE sensors, preamplifiers (if applicable), signal processor (particularly the signal processor time reference), and AE electronic simulator (waveform generator) should be performed. Equipment should conform to manufacturers specifications. Instruments should be calibrated in accordance with National Institute for Standards and Technology (NIST) specifications. An AE electronic simulator, used in making evaluations, must have each channel respond with a peak amplitude reading within 2dBV of the electronic waveform output. A system performance check should be done immediately before and after an AE examination. The preferred technique for this is the pencil lead break test. A complete description of this test can be found in ASTM E 570. One very important factor affecting sensor performance is location. Determination of the number of sensors required for the test, their placement strategy and location on the specimen to be monitored is critical. A single sensor used near the expected source of AE is sufficient when background noise can be controlled or does not exist. When background noise is limited, the use of a single AE data sensor near the expected source plus a guard sensor(s) near the background source will suffice. ASTM defines a guard sensor as sensors whose primary function is the elimination of extraneous noise based on arrival sequences. Another technique involves the placement of two or more sensors to perform spatial discrimination of background noise and allow AE events to occur. ASTM defines spatial discrimination as the process of using one or more (guard and data) sensors to eliminate extraneous noise based on arrival sequences.

PAGE 55

42 In situations where irrelevant noise cannot be controlled during testing and could be emanating from any and all directions, a multiple sensor location strategy should be considered. Using a linear or planar sensor configuration will allow for accurate source location of the acoustic emission event. Applications of spatial filtering and/or spatial discrimination will only allow data emanating from the region of interest to be processed as relevant AE data. Preamplifiers Preamplifiers (as shown in Figure 4.3) are used to prevent loss of sensor activity. Loss occurs when one sensor is connected through a long coaxial cable to an amplifier. The amplifier is split into a fixed gain preamplifier located close to the sensor. The preamplifier consists of a low noise input stage, bandpass filters and a low impedance output stage capable of driving a 50-ohm cable (Beattie, 1993). Power for the preamplifier is received from the main instrument group. AE preamplifiers are designed to have a relatively flat frequency response between about 20 kHz and 2MHz, without the bandpass filters (Beattie, 1993). Preamplifiers can be included in the sensor package. The advantages of this arrangement are the elimination of cable capacity effect and being able to tailor the preamplifier characteristics to match the sensor. The disadvantages of such units, besides their higher cost, are that they are restricted to temperatures near 20C (the preamplifier will not work properly at higher or lower temperatures) and that a separate preamplifier has to be purchased for each sensor (Beattie, 1993). Postamplifiers and Signal Processors Most AE systems use variable gain postamplifiers. This allows the use of signal processors with fixed input ranges or thresholds in conjunction with fixed gain

PAGE 56

43 preamplifiers (Beattie, 1993). The systems total gain is the sum of the preamplifier and postamplifier gains, expressed in decibels. Additional noise reduction can be achieved through postamplifiers from bandpass filters. Signal processors are typically included in the systems capabilities. These include voltage controlled gates that allow data to be collected only on certain portions of a load cycle, envelope processors which attempt to filter out high frequencies leaving only the signal envelope to be counted, logarithmic converters which allow the output of the signal analyzer electronics to be plotted in logarithmic form and a unit which allows the combination of outputs from several preamplifiers so that several sensors can be monitored by one channel of electronics (Beattie, 1993). Figure 4.3: Pre-amplifiers (PAC) Transient Recorders Transient recorders are used to study individual AE burst signals. A signal is digitized in real time, and then stored into memory. A transient recorder is used in sequence with an oscilloscope or spectrum analyzer to display AE signals at visible speeds. Digital rates vary on transient recorders. The fastest rate of the recorder is ultimately the limiting rate, with some instruments sampling up to 1 word/nano-second (Beattie, 1993).

PAGE 57

44 Sampling rates can be modified for testing purposes. One advantage of transient recorders is an additional mode of triggering and pretriggering, where the input signal is continuously digitized and the data stored in the memory (Beattie, 1993). This feature allows a digitized picture of the signal to be displayed as it is received. More advanced systems allow recording of two or more signals simultaneously. The recording of more than one AE signal is shown in Figure 4.4. Figure 4.4: Transient recorder with multiple AE signals Spectrum Analyzers The ideal spectrum analyzer has horizontal and vertical signal outputs, allowing an X-Y plot of the AE signal. Spectrum analysis is done either directly with analog electronics or by digitally viewing a continuous signal. A local oscillator signal is mixed with the input signal and the highest frequency is passed through a chain of intermediate frequency (I.F.) amplifiers, after which it is measured by a voltmeter (Beattie, 1993). The oscillator is swept through a frequency range so that the frequency components of the signal selected by the I.F. amplifier are continuously changing (Beattie, 1993). The

PAGE 58

45 voltmeter output is plotted on the vertical axis of the oscilloscope. A typical oscilloscope is shown in Figure 4.5. A synchronized signal is displayed for the horizontal axis through the local oscillator frequency. The final result is a plot of signal strength vs. frequency. Spectrum analyzers range in frequencies from at least 10kHz to 2MHz. The width and speed of the local oscillator and the sharpness of the I.F. amplifier filters are all under the direct control of the operator (Beattie, 1993). An AE burst signal emission is not suitable for spectrum analysis. AE burst signals must be captured on a tape recorder and played repetitively onto the spectrum analyzer for analysis. This essentially simulates a continuous signal for spectrum analysis. Figure 4.5: Digital Oscilloscope LAM--Local Area Monitor LAM is the worlds first acoustic emissions system to allow remote condition monitoring of structures. Physical Acoustics Corporation developed the LAM, as shown in Figure 4.6, in conjunction with the U.S. Federal Highway Administration. The system

PAGE 59

46 is portable and easy to handle weighing only 25 pounds including one battery pack. Features of the LAM used in this research project are: Modular 8 channel DSP-based AE system with 16-bit A/D User-friendly software AC/DC powered 4 high-speed and 8 low-speed parametrics Digital AE features and waveforms processed simultaneously Software programmable filters Resistant to harsh environmental conditions Figure 4.6: LAM-Local Area Monitor The LAM is useful for monitoring many structures including bridges for defects, chemical/ petrochemical tanks for leaks and deterioration, transformers for partial discharge and most structures (including concrete pipes) during fatigue tests. The LAM was originally designed for monitoring defects in steel bridges. The LAMs modularity lends itself to many other applications, as stated before, to monitor fatigue cracks and other discontinuities in structures, pressure vessels and transformers. This research will use the LAM to monitor microcracking in the walls of fiber reinforced and steel reinforced concrete pipes.

PAGE 60

47 The LAM offers up to 8 channels of digital AE for short-term condition monitoring, long-term integrity monitoring, laboratory fatigue testing or incipient failure detection monitoring through user selections. The unit operates from an external 12 Volt DC battery or 110 Volt AC power supply. An optional feature is remote access through traditional phone line or cellular phone. This feature allows the user to monitor the apparatus from an office or other location remote from the test site. Advantages of the LAM over other acoustic emission instrumentation include a reduction in the extensive cabling normally required for operation. The unit is placed on site with the structure/ object to be monitored. For this project, the concrete pipes will be instrumented with the eight sensors to monitor microcracks occurring inside the walls. Data analysis for this research will be performed with the NOESIS 3.1 software published by Physical Acoustics Corporation. NOESIS 3.1 is an MS Windows based advanced data analysis pattern recognition and neural networks software used for acoustic emission applications. It provides all necessary tools for analyzing, filtering and classifying acoustic emission hits and waveforms that are acquired with the LAM unit. NOESIS is equipped to handle data saved in a DTA file format produced by the LAM unit. The software utilizes PAC (Physical Acoustics Corporation) file libraries to load and save data in the DTA file format. NOESIS allows multiple DTA files to be loaded simultaneously for direct comparison, statistical analysis and filtering or merging. Direct export of data files to MS EXCEL and MS WORD is also possible. Any number of windows can be displayed, limited only by the resolution of the viewing window for adequate visibility. Navigation

PAGE 61

48 throughout the program, along with data selection, is done with the mouse. Data point selection is available using the mouse on scatter plots, cumulative plots, waveform plots, FFT plots and tabular data views. NOESIS allows graphs and plots to be customized for presentation of data and analysis. Waveforms and AE hits can be selected and displayed on any graph or data table. Comparison of the data can be superimposed or viewed in three dimensions. Graphical and other data filtering are applied to waveform views for presenting the collected data. Any changes made to hardware settings are immediately reflected in the waveforms. Minolta 3D Digitizer--VIVID 900 Minolta is the worlds largest manufacturer of 3-D non-contact digitizing instruments, providing a 3-D scanner with a simple point and shoot camera that scans 300,000 points in less than 3 seconds. This project will use Minoltas newest 3-D scanner, the VIVID 900, shown in figure 4.7. The VIVID 900 is an easy to use scanner with simplicity, flexibility and portability. Minolta offers simplicity by a point and shoot camera with excellent results. The VIVID 900 includes interchangeable lenses, for variable scanning volumes, for flexibility. The camera unit is compact, measuring 8-3/8x16-1/4x 10-11/16 and weighs only 24 lbs. Scans can be saved and stored on a compact flash memory card or viewed immediately after scanning on the rear-panels color LCD viewfinder. Color images are equivalent to a 3 CCD digital camera displaying full 24-bit color depth. Hardware The Minolta VIVID 900 offers variable volumes for digitizing between 110 x 80 x 40 mm and 1200 x 900 x 750 mm. There are three interchangeable lenses included as standard accessories for scanning; telephoto, medium and wide angle. The VIVID 900 is

PAGE 62

49 an independent instrument that does not require a host computer for operation. Scanned images are saved to a flash memory card (Figure 4.8) or viewed immediately after scanning on the LCD viewfinder. The Minolta VIVID 900 also offers an autofocus function that eliminates the need to move the unit back and forth to achieve optimal focus for the scan. Figure 4.7 Minolta VIVID 900 Non-Contact 3-D Digitizer Figure 4.8 Compact Flash Memory Card (40MB capacity) Accessories Minolta offers accessories to improve the outcome of the scan. When scanning a complete 360 view of an object, the rotary specimen stage shown in Figure 4.9 will

PAGE 63

50 prove helpful. A rotating disc is set to rotate at a specified speed while the VIVID 900 scans the object, resulting in a full 3-D image of the object. To ensure level scans and stability, Minolta offers a tripod accessory with the option of a tilt mounting base unit as shown in figure 4.10. Other accessories available are a PC card adapter allowing direct transfer of the scanned images to a personal computer for analysis. Figure 4.9 Rotating Stage Set for Scanning a Full 3-D image Figure 4.10 Tripod (left) and Tilting Base Mount (right) for Minolta VIVID 900

PAGE 64

51 Operation The Minolta VIVID 900s basic theory of operation is described through LASER triangulation. A laser source from the VIVID 900 emits a horizontal light stripe through a cylindrical lens onto the object being scanned. The plane of light is swept across the field of view by a rotating mirror. The light is reflected from the scanned object, captured, and observed by a single frame through the CCD camera. Once received by the CCD, the light is converted through triangulation into distance information. Each scan line is captured and observed by the CCD camera. The shape of the image of each reflected scan line is derived to produce the contour of the surface. The selected area in the view is captured in 2.5 seconds (0.3 seconds in FAST mode), and the surface shape is converted to a lattice of over 300,000 vertices or connected points (Minolta, 2001). The VIVID 900 produces polygonal-mesh with all connected information, eliminating geometric ambiguities while improving detail. Minoltas VIVID 900 uses an X, Y, and Z coordinate axis. The x coordinate is the horizontal dimension of the focal plane, the y coordinate is the vertical axis and the z coordinate is the distance from the sensor. The VIVID 900 creates no parallax error. Specifications for the VIVID 900 are displayed in Table 4.1 below.

PAGE 65

52 Table 4.1 Specifications of Minolta VIVID 900 3D digitizer. Type Non-contact 3D digitizer VIVID 900 Measuring method Triangulation light block method AF Image surface AF (contrast method), active AF Light-Receiving Lens (Exchangeable) TELE: Focal distance f=25mm MIDDLE: Focal distance f=14mm WIDE: Focal distance f=8mm Image Input Range 0.6 to 2.5m (2m for WIDE) Measurement Input Range 0.6 to 1.2m Laser Output Eye-safe, Class I (FDA), Class 2 (IEC), Maximum 30mW 690 nm Laser Scan Method Galvano mirror Input Time 0.3 sec (FAST mode), 2.5 sec (FINE mode), 0.5 sec (FINE mode) Transfer Time to Host Computer Approx. 1 sec (FAST mode, 1.5 sec (FINE mode) Ambient Light Condition Office Environment, 500 1x or less Imaging Element 3-D data: 1/3-inch frame transfer CCD (340,000 pixels) Color data: 3-D data is shared (color separation by rotary filter). Number of Output Pixels 3-D data: 640 x 480 (for FINE mode); 320 x 240 (for FAST mode) Color data: 640 x 480 Output Format 3-D data: Minolta format, & (STL, DXF, OBJ, ASCII points, VRML) (Converted to 3-D data by the Polygon Editing Software/ standard accessory) Color data: RGB 24-bit raster scan data Recording Medium Compact Flash memory card Data File Size Total 3-D and color data capacity: 1.6MB per data (for FAST mode), 3.6MB per data (for FINE mode) Viewfinder 5.7-inch LCD (320 x 240 pixels) Output Interface SCSI II (DMA synchronous transfer) Power Commercial AC power100 to 240V (50 to 60Hz), Rated current 0.6A (when 100Vac is input) Dimensions (WxHxD) 213 x 413 x 271 mm (8-3/8 x 16-1/4 x 10-11/16 in.) Weight Approx. 11 kg. Operating environment Temperature: 10-40C (50-104F); relative humidity 65% or less with no condensation, Pollution degree:2, Installation category:II Storage Temperature -10 to 50C (14-122F); relative humidity 85% or less (at 35C/95F) with no condensation

PAGE 66

CHAPTER 5 PLAXIS VERSION 7.2--FINITE ELEMENT CODE FOR SOIL AND ROCK ANALYSIS Introduction Plaxis was developed in 1987 at the Technical University of Delft, initiated by the Dutch Department of Public Works and Water Management. Initially, Plaxis was developed to analyze river embankments in the soft soils of the Dutch lowlands. Throughout the original development and later improvements, Plaxis extended its applicability to cover most areas of geotechnical engineering. Plaxis excelled over the years forming a company in 1993, Plaxis BV. Plaxis is a computer program designed to provide a practical analysis tool for use by engineers who are not specialists in finite element analysis. Non-linear finite element computations done without a computer can be too time-consuming for regular analyses. Plaxis 7.2 is an MS Windows-based program with easy-to-use tabs used to navigate through the analysis. Plaxis is a finite element program designed for the analysis of deformation and stability in geotechnical engineering projects. Geotechnical engineering uses advanced models for the simulation of non-linear and time dependent behavior of soils. Soil is a multiphase material with properties that can change with a changing environment. Plaxis is equipped to deal with hydrostatic and non-hydrostatic pore pressure in the soil. Even though modeling of the soil itself is important, modeling of soil-structure interactions is 53

PAGE 67

54 the application seen in many engineering projects today. Plaxis is also equipped with features to analyze a number of aspects dealing with complex geotechnical structures. Plaxis is made up of four internal programs, Input, Calculation, Output and Curves. Graphical input of geometry models consists of soil layers, structures, construction stages, loads and boundary conditions all created with drafting procedures on a CAD (Computer Aided Drawing) screen, as shown in Figure 5.1. The CAD interface allows accuracy and detailed modeling of real engineering situations. From the geometry entered, a finite element mesh is generated. The automatic mesh generation feature allows for fully automated mesh generation of unstructured finite element meshes. Figure 5.1: Plaxis 7.2 Computer Aided Drafting screen used to create modeling analysis. Beam elements are used to model retaining walls, tunnel linings and other structures. Pipes are modeled using the tunnel feature. Behavior of each element is defined as flexural rigidity, normal stiffness and/or ultimate bending moment. Assigned

PAGE 68

55 to the beam elements is a feature called interface. Interfaces are joint elements used in calculations of soil-structure interactions. For example, interfaces can simulate a zone of intense shearing stresses located at the contact plane of footings, piles, geotextiles, retaining walls and pipe surfaces. Values of friction angle and adhesion properties can be assigned to interface elements. Anchors and geotextiles are additional features available for modeling. The tunnel feature option creates circular and non circular tunnels composed of arcs. Beams and interfaces may be added for analysis of tunnel linings or interactions with the surrounding soil. A number of soil models are used in Plaxis for analysis of soil performance. Mohr-Coulomb, a simple non-linear model, is the most used model based on soil parameters encountered in everyday practical situations. Other advanced soil models are also available for analyses. Pore pressures are analyzed within the Plaxis model. Steady state and excess pore pressures are defined by the water table location in the model. There are two approaches for steady state pore pressures. Complex pore pressure distributions are generated on the basis of a two-dimensional groundwater flow analysis. For simple conditions, multi-linear pore pressure distributions can be directly generated by a simple phreatic line in the model. Excess pore pressures are computed during plastic calculations when pores are full of water and subjected to loads. A typical analysis with Plaxis finite element modeling involves input parameters and specifications of model types. The input model is then run through the calculation phase to produce output results for analysis. Plaxis is also equipped with a curves

PAGE 69

56 program that produces graphs for analysis results, such as stress/strain curves and load/displacement curves. Plaxis--Input The input program is the first step of a Plaxis analysis. At the start of a new project, the General Settings window appears (as shown in Figure 5.2), prompting the user to set the basic parameters of the project. The General Settings window has two tabs: Project and Dimensions. The project tab allows the user to enter a description of the project, the type of model used (i.e. plane strain) and the number of nodes to be used in the finite element analysis (i.e., 6 and 15 node). The Dimensions Tab specifies units of length (ft), force (lbs.) and time (day); and the geometry dimensions for the CAD screen and grid spacing. Figure 5.2: General Settings window in Plaxis 7.2.

PAGE 70

57 Once the general setting parameters are entered into Plaxis, the CAD screen appears for geometric creation of the model. An example of a tunnel model, as used in this research, will be used to demonstrate a typical run-through of a Plaxis 7.2 finite element modeling. The tunnel model can be used to simulate buried pipes for analyses. The model described in the example is a four-foot diameter reinforced concrete pipe, located within a soil box. Ultimately, a steel framed box will be designed to test concrete pipes. Plaxis modeling will be performed on the soil structure interaction and in analyzing the stresses induced on the sidewalls and pipe, thus defining the design parameter for the box. After the general settings are complete, the geometry of the model is created in the input drawing area using the points and lines feature. With the Windows based program the creation of a model is accomplished by working from left to right with the icons at the top of the screen, as shown in Figure 5.3. Points and lines are used to create an enclosed box of dimensions 20 x 10 x 10. The tunnel option is then used to create a circular tunnel representing the four-foot diameter concrete pipe, composed of arcs defined by a radius and a radial increment (angle). By clicking on the tunnel feature, the user is prompted with a choice to use a whole tunnel or a half tunnel. After this choice is made, a window is opened as shown in Figure 5.4 to allow input of the radius and radial increment. In this same window an interface and lining material is assigned to the tunnel model. For the example, a radius of 4 feet is entered and a tunnel and interface lining is assigned to the tunnel. An interface is used to model soil-structure interaction and the

PAGE 71

58 lining material provides the option to assign concrete properties to the tunnel lining. The next step in the model is load and boundary conditions. Figure 5.3: Plaxis 7.2 Main Toolbar. Figure 5.4: Tunnel Designer in Plaxis 7.2. The load and boundary conditions used for this model are standard fixities and traction loads (distributed loads). Both features are chosen from the standard toolbar as shown in Figure 5.3. The soil box will be modeled as a rigid box with minimal deflection. Fixities are prescribed displacements equal to zero. These fixities can be applied to points or lines. By clicking on the standard fixities button on the toolbar, the two sidewalls and the bottom of the box were assigned both horizontal and vertical fixities as shown in Figure 5.5 below.

PAGE 72

59 After the standard fixities are assigned, moving to the left on the toolbar, the loads are assigned through the traction icon as shown in Figure 5.5. A distributed load was assigned using traction A, assigning the load to the two top points of the box on the CAD drawing area. Plaxis will ask the user to enter a multiplier for loading steps in the calculation phase. The default is 1.00, representing a true value for the load entered in the calculation phase. A multiplier of 1.00 was used in the example. Figure 5.5: Standard Fixities in Plaxis 7.2 shown on a soil box with right half tunnel. Once geometric definition of the model is finished, with all beam elements defined, fixities assigned and loads designated, the material properties need to specified and added to the model. Material data sets for soil, interfaces and beams are entered using the material data set icon located on the main toolbar. When the icon is clicked, a Material Sets window will open with choices for the project database as shown in Figure 5.6.

PAGE 73

60 The example requires material data sets for soil & interfaces and beams. The first step is to set the parameters for the soil & interfaces. Input parameters for the soil are entered in the window shown in Figure 5.7 and include material model, unit weight, permeability, stiffness, poissons ratio, cohesion, friction and dilatancy angles and interface properties. The input parameters used for the example soil model are as follows: Model used: Mohr Coulomb Material type: drained Unit weight: 120 pcf Permeability: 0 (both vertical and horizontal) Stiffness (E): 489,600 psf Poissons ratio (): 0.3 Cohesion (c): 1 x 10 -5 Phi angle (): 35.00 Dilatancy angle (): 5.00 Specified along with the input parameters for the soil model are the settings for the interface. The soil interface is specified as either rigid or set manually. This feature is used for soil-structure modeling. The example model shown is modeling the soil-structure interaction between the soil and a concrete pipe. Both rigid and manual settings for the interface are used. Rigid assigns the soil properties to the interface and Manual allows the user to specify a friction angle for modeling two materials in contact with one another, such as concrete and soil or steel and soil.

PAGE 74

61 Figure 5.6: Material Sets Window in Plaxis 7.2. Figure 5.7: Soil Input in Plaxis 7.2 Mohr Coulomb Model. The material set for beams is specified when the user selects beams from the material set window and assigns the parameters. Input parameters are entered in the

PAGE 75

62 window as shown in Figure 5.8. The input parameters used for the beam material model representing the concrete pipe are: Axial Stiffness (EA): 107.3 x 10 6 lb/ft Flexural Rigidity (EI): 193.5 x 10 6 lb-ft/ft Unit weight: 150 pcf Poissons ratio: 0.15 The axial stiffness and the flexural rigidity are calculated using the modulus of elasticity and the cross sectional area of the concrete pipe along with the moment of inertia of the pipe. An estimate for the unit weight was obtained from a concrete manual. Figure 5.8: Beam Properties Input Window in Plaxis 7.2. After the material parameters are assigned to the model, the material sets window appears once more to click and drag the material properties onto the model. Once this task is complete the window is closed. At this point, the model setup is complete. In order for the finite element analysis to proceed, however, the geometry must be divided into elements called a finite element mesh. The mesh generation icon is the last icon on the main tool bar, located to the far right. Plaxis is equipped with a self

PAGE 76

63 generating mesh tool. Once the mesh is generated, a plot is displayed through the output program. The mesh can be defined as very coarse, coarse, medium, fine and very fine. A finer mesh is used where large stress contributions might be seen due to loading. After the mesh is generated the initial conditions are entered as shown in Figure 5.9. Initial conditions involve an initial stress state before loading and an initial situation. This process is still part of the input program. The initial conditions consist of two different modes; one for the generation of initial water pressures (water conditions mode) and one mode for the specification of the initial geometry configuration and the generation of an initial effective stress field (geometry configuration mode). Switching between these two modes is done by clicking on icons as shown in Figure 5.9. Figure 5.9: Pore Water Pressure & Initial Stress Modes in Plaxis 7.2. Water conditions are specified by means of the water weight and phreatic lines. The model used for this example was in a dry state, thus there were no phreatic lines or water pressures to generate. The screen is then switched to initial geometry configuration. In this step the example model is shown with the soil box and concrete pipe inside the box. The initial geometry configuration allows the user to select geometry objects that are not active in

PAGE 77

64 the initial situation. This means that the beam elements can be turned off to determine the initial stresses before addition of the concrete pipe to the soil box. The initial stresses are calculated using the K 0 procedure with a default value of 0.426 when the stress icon is chosen on the tool bar. An initial vertical stress is calculated using the coefficient of lateral earth pressure K 0 After the initial stresses are generated, a window is displayed with the initial stresses showing the plane of direction. Once the input stage is completed, the next step is the calculation stage. Plaxis--Calculations After the mesh is generated, the finite element calculations can be specified and executed in the calculations program. In this program each type of calculation to be performed is defined, along with the type of loading activated during the calculation. Initiating the calculation program will open a window as shown in Figure 5.10. The three tabs (general, parameters and multipliers) will allow the user to navigate through to the end calculate command. The only settings to define in the calculation program are the calculation type, selecting the calculation phases and multipliers. The other input areas in the calculations program were set to default and were used as default for the example model.

PAGE 78

65 Figure 5.10: Plaxis 7.2 Calculations Program. In the calculation program under the general tab the calculation type is specified along with the setup of the calculation phase(s) using the insert button as seen in Figure 5.10. The insert button will add additional phases for calculation. In the example model, three phases were setup; initial phase, phase 1 and phase 2. Phases are assigned parameters and multipliers when highlighted in the phase display box located at the bottom of the calculations window. The initial phase represents the starting conditions of the project as defined in the initial conditions mode of the Input program. For the example model, two phases were added: staged construction and total multipliers. The staged construction represents the installation of the concrete pipe with loading coming from overburden soil above the pipe. The total multipliers stage is where the assigned distributed load is activated and applied to the model. Under the parameters tab (as shown in Figure 5.11) the default modes are used. The parameters tab is where the staged construction phase is defined. The loading input group is used to specify which type of loading is considered in a particular calculation

PAGE 79

66 phase. Loading input was set to stage construction. The define button, located at the bottom right, will open the input window of the model allowing the user to deactivate and activate soil clusters and structural objects that define the construction of the model. After the staged construction phase is defined, the update key will return the user to the calculation program. The multipliers tab is where loads are assigned for the total multipliers stage. There are two types of multipliers: incremental and total as shown in Figure 5.12. In the example model shown, the total multiplier assigned in the input program is MloadA at a value of 16,000 with units of pounds per square foot. MloadA controls the magnitude of the traction loads as defined in the Load System A of the input program. Figure 5.11: Plaxis 7.2 Calculations Program Parameters Tab

PAGE 80

67 Figure 5.12: Plaxis Calculations Program Multipliers Tab Another feature in the calculation phase sets up points for curves generated in the curves program. The points can be entered by selecting the Select Points For Curves option from the View menu or by clicking on the corresponding button in the tool bar. Selection occurs when the output program opens showing a plot of the finite element mesh displaying all of the nodes. Nodes are selected by left clicking the mouse on the node of interest. Each node selected is characterized in the curves program by an alphabetical letter. A node can also be deselected by clicking on it again with the mouse. In the example model shown, points were selected to create curves for load displacement and stress/strain curves. After selecting the points for curves the calculate button will run the calculation program. A window is opened to view the loading increments displaying different properties. Once the calculation is completed the calculation window appears with green checks beside the phases of calculation. Output of the calculation is viewed with the output program, which is opened by clicking on the output button located at the top of the calculation window.

PAGE 81

68 Plaxis--Output The main output quantities of a finite element calculation are the displacements at the nodes and the stresses at the stress points. When a finite element model involves structural elements, structural forces are calculated in these elements. The output program is equipped to display the results of a finite element analysis. The resulting program window is displayed in Figure 5.13, showing the deformed finite element mesh due to the load. Results are viewed via the output program, displaying deformations, deformed mesh, total displacements, total increments, total strains, incremental strains, stresses, effective stresses, total stresses, plastic points, active pore pressures, excess pore pressures, groundwater head, flow field, structures and interfaces, beams, geotextiles, interfaces and anchors. In the example model (a four foot diameter concrete pipe modeled in a 20 x 10 x 10 soil box) the only outputs of interest are deformed mesh, total displacements (how much the pipe and soil settled), stresses, structures and interfaces and beams (concrete pipe). Each of the output results can be viewed as a picture, table or curve. The first plot the user views in the output program is the deformed mesh plot as seen in Figure 5.13. From this plot the user can navigate to other results by selecting from the menu at the top of the output window. In the example model, the first thing to analyze is the stress distribution throughout the soil contained in the box. Figure 5.14 displays the stress distribution throughout the soil box by means of shading, with red indicating the highest stresses. Shading provides a colorful way of presenting results for effective mean stresses. Another way to display stress distribution results is through the contour plot of effective mean stresses as shown in Figure 5.15. The contour plot

PAGE 82

69 displays a legend on the right side of the window that identifies each contour. A scan line, labels the contours with respect to the legend. Figure 5.13 Plaxis output program with deformed mesh displayed on example model. Figure 5.14: Plaxis Output Effective Mean Stresses Displayed by Mean Shading.

PAGE 83

70 Figure 5.15: Plaxis Output Effective Mean Stresses Displayed by Contours. Cross sections through the model can be used to present results of displacement and stress distributions. Viewing output in a cross section allows the user to gain insight into the distribution of a certain quantities calculated by the model. Cross sections are created in the model by selecting the cross section button and then clicking and dragging a line through the model where desired. In the example model stress distribution on the sidewalls was desired, requiring a plot of stress distribution along the sidewalls of the soil box. The cross section at the sidewall is shown in Figure 5.16. Also shown with the cross section tool are the horizontal displacements along the cross section shown in Figure 5.17. Beam elements, as created by the input program, are viewed in the output program with element and interface results. Element results are the beam forces, displacements and bending moments. Interfaces assigned to the element will show displacements and stresses acting at or within that interface. Displacement output for the beam element

PAGE 84

71 concrete pipe and the interface assigned to the concrete pipe are shown in Figure 5.18. Plaxis will specify the extreme displacement for the beam elements and the interfaces. The bending moment of the concrete pipe in the example model is shown in Figure 5.19. Beam element properties are displayed in the output program by double clicking on the element itself. The various display choices are activated from the sub menu found at the top of the screen. Figure 5.16: Stress Distribution Cross Section A-A in Plaxis 7.2 Output Program

PAGE 85

72 Figure 5.17: Horizontal Displacement Cross Section A-A in Plaxis 7.2 Output Program Figure 5.18: Displacements for the Pipe and Interface Plaxis 7.2 Output Program.

PAGE 86

73 Figure 5.19: Bending Moment for the Pipe Plaxis 7.2 Output Program. The output program is also capable of producing tabular data for analysis. The numerical data can be viewed in output tables for all types of graphical output by clicking on the Table button in the main tool bar or by selecting the Table option from the View menu. A menu is available to view selections of other quantities. Tables available in Plaxis include displacement, stresses and strains, and stresses and forces. In the research example used, tabular data was not used directly. Output data was fed into the curves program to allow generation of different graphs.

PAGE 87

CHAPTER 6 FINITE ELEMENT ANALYSIS OF A SOIL BOX TEST FACILITY Introduction A finite element analysis was performed in order to provide the necessary information for the design and fabrication of a soil box for testing two different concrete pipes, fiber reinforced and standard reinforced. Two and three-dimensional finite element analyses were done in order to evaluate the stress on the sidewalls due to the applied load and the effect of boundary conditions (i.e., friction on the sidewalls and around the perimeter of the test pipe). The soil box will be fabricated of steel and filled with compacted soil similar to that used by the State of Florida in highway right-of-ways. The two-dimensional program used, Plaxis 2D, is a finite element analysis for soil and rock. Plaxis 2D evaluated the stresses on the sidewalls and the soil structure interaction between the soil and concrete pipes as well as between the soil and steel structure sidewalls. From the two dimensional analysis, the box dimensions were examined and the stresses evaluated using an interface between the soil and concrete pipe and the soil and the sidewalls. In the analysis, boundary conditions were evaluated in order to determine size requirements that would not induce lateral stresses on the test pipe. The results from the two-dimensional analysis are included in Appendix A. A three-dimensional finite element analysis, Plaxis 3D Tunnel was used to evaluate the two-dimensional analysis, plus stresses in the z direction, the third dimension. Plaxis 3D analyzed the stress distribution throughout the box while varying the friction angle at 74

PAGE 88

75 the soil-sidewall and soil-pipe interfaces to examine the stress concentrations induced within the soil box. Each of the three possible lengths to be considered (10, 15, and 20) was evaluated while varying the friction angle and analyzing the boundary conditions. Displacements of the concrete pipes were evaluated with and without friction on the sidewalls. Stresses were examined along the length of the pipe to ensure that no shear stresses were induced on the ends of the pipes near the front and rear walls. Two different manufactured pipes will be tested inside the soil box, fiber reinforced and standard reinforced concrete pipes. The proposed pipe diameters to be used for testing are 18 and 24 with the further possibility of sizes as large as 48 in diameter. In order to ensure proper bedding, a depth of at least one diameter below the pipe inside the soil box will be used in the analysis. For example, the 24 diameter pipe will need two feet of underlying soil, resulting in an approximate height of six feet above the crown of the pipe to the top of the box. An overburden soil depth of six feet will allow a distribution of stresses that simulates in-situ conditions. The maximum load applied on the soil will be 16,000 lbs/ft 2 The two types of backfill used in the analysis (shown in Table 6.1) were loose and dense compacted soil. For the loose compacted soil, a Youngs Modulus of 216,000 lbs/ft 2 was used, while the dense compacted soil had a Youngs Modulus of 489,600 lbs/ft 2 Input parameters for the different sizes of SRCP and FRCP are displayed in Tables 6.2-6.4. Youngs Modulus for the FRCP and SRCP were obtained from previously published literature. The SRCP modulus value was calculated using the American Concrete Institute (ACI 318) relationship between elastic modulus and compressive

PAGE 89

76 strength, as shown in Equation 6.1. Compressive strength for concrete pipe normally ranges from 4000 lbs/in 2 to 6000 lbs/in 2 (Rinker Materials, 2003). Elastic modulus for the SRCP was calculated using a compressive strength (f c ) of 4000 lbs/in 2 The elastic modulus for the FRCP, used to determine normal stiffness and flexural rigidity was 25 Gpa or 3.62 x 10 6 lbs/in 2 ccfE'*000,57 (6.1) Table 6.1 Material Properties of the soil (Loose & Dense). Parameter Name Loose Dense Unit Material Model Model Mohr-Coulomb Mohr-Coulomb Type of material behavior Type Drained Drained Soil weight above phr. level unsat 120 120 lbs/ft 3 Soil weight below phr. level sat 120 120 lbs/ft 3 Youngs modulus E ref 216,000 489,600 lbs/ft 3 Poissons ratio 0.3 0.3 Cohesion c ref 0.00001 0.00001 lbs/ft 3 Friction angle 35 35 DEG. Dilatancy angle 5 5 DEG. Table 6.2 Material Properties of the 18 diameter concrete pipes (FRCP & SRCP). Parameter Name FRCP SRCP Unit Type of behavior Material type Elastic Elastic Normal stiffness EA 147,372,845 280,303,765 lbs/ft Flexural Rigidity EI 38,138,787 66791264 lbs-ft 2 /ft Weight W 150 150 lbs/ft 2 Poissons ratio 0.15 0.15 Table 6.3 Material Properties of the 24 diameter concrete pipes (FRCP & SRCP). Parameter Name FRCP SRCP Unit Type of behavior Material type Elastic Elastic Normal stiffness EA 198,636,143 382,227,687 lbs/ft Flexural Rigidity EI 93,303,588 168,720,067 lbs-ft 2 /ft Weight W 150 150 lbs/ft 2 Poissons ratio 0.15 0.15

PAGE 90

77 Table 6.4 Material Properties of the 48 diameter concrete pipes (FRCP & SRCP). Parameter Name FRCP SRCP Unit Type of behavior Material type Elastic Elastic Normal stiffness EA 665,723,216 1,228,267,715 lbs/ft Flexural Rigidity EI 1,263,829,838 2,322,073,435 lbs-ft 2 /ft Weight W 150 150 lbs/ft 2 Poissons ratio 0.15 0.15 Plaxis 3D--Verification Analysis A three-dimensional analysis was performed to verify the results obtained from the two-dimensional analysis, as well as to analyze the stresses and displacements in the third dimension. Plaxis 3D introduces a third dimension of analysis allowing stresses in the z-direction to be examined. From the two-dimensional analysis, results showed that the use of well-compacted soil (90% proctor) provided proper stability with half as much settlement as that of the loose compacted soil. It is important to obtain proper compaction during the installation of the pipe to produce maximum performance of the pipe. Three-dimensional finite element analysis was used to verify eight analyses performed with the two-dimensional finite element program (Appendix A). Verification of the stresses and displacements throughout the soil box were done for the 18 and the 24 diameter FRCP and SRCP. The ability to test a 48 diameter pipe in the soil box is questionable according to the two-dimensional analysis due to the limited depth of soil cover on top of the test pipe. Verification of the 18 and 24 diameter pipe was done using the three dimensional version of Plaxis. Four different analyses of the 18 and 24 pipe were performed using dense compacted soil and varying the friction angle along the wall within the interface for each of the SRCP and FRCP. The three-dimensional analyses differed very little from

PAGE 91

78 the two-dimensional analyses, through the three-dimensional results provided better stress distribution throughout the soil box in the third dimension. Varying the soil-wall friction angle from a value equal to the soil friction angle of 35 degrees to less than 5 degrees doubled the amount of total displacements throughout the whole soil box. Assigning a value of 5 degrees to the interface resulted in higher displacements within the soil upon loading. Friction on the sidewalls equal to that of the soil resulted in a total displacement of 0.236 while a friction angle of less than 5 degrees on the sidewalls results in a total displacement of 0.433. When comparing the 18 and 24 diameter pipes, the total displacements were the same. Figure 6.1 shows the total displacements for the 24 FRCP and Figure 6.2 shows the total displacement for the 24 SRCP, both with a friction angle equal to that of the soil. It is visually apparent on the sidewalls that there is friction. The load is distributed throughout the third dimension and the friction along the wall is visible through the shading. The effective mean stress reported from the three-dimensional analysis differed little from the two-dimensional analysis. For example, the extreme effective mean stress for the 18 SRCP, friction sidewalls, measured 12,690 lbs/ft 2 in the two-dimensional and 12,520 lbs/ft 2 in the three-dimensional analysis. The three-dimensional analysis distributes the stresses more efficiently in the z direction (i.e., the third dimension). The analysis of the box focuses on the maximum stress on the sidewalls as a result of the 16,000-lbs/ft 2 load. Each of the two-dimensional wall friction analyses examined the effective mean stress on the sidewalls. When verified by the three-dimensional finite

PAGE 92

79 Figure 6.1 A Three Dimensional View of Total Displacements for 24 FRCP. Figure 6.2 Three Dimensional View of Total Displacements for 24 SRCP

PAGE 93

80 element program, it was found that the sidewall stresses were slightly smaller. Sidewall stresses examined through the interface along the sidewall increased when the friction angle was set to 35 degrees. For example, the interface wall stress for the 18 diameter SRCP with friction was 7,070-lbs/ft 2 compared to an interface without friction of 4,470-lbs/ft 2 The soil was unable to move freely upon load application when friction occurred at the sidewalls, therefore creating higher stresses on the sidewalls. Figure 6.3 shows the left sidewall of the soil box displaying the effective normal stresses. Figure 6.3 Left Side Interface of Soil Box Model 24 FRCP An interface was also assigned to the perimeter of the pipe inside the soil box for soil-structure interaction modeling. The friction angle associated with this interface was also varied in the same manner as the sidewalls. The extreme effective mean stress

PAGE 94

81 around the pipe and the extreme total displacement around the pipe are reported for each of the tests. There was very little difference between the FRCP and the SRCP modeled. In varying the friction angle at the interface around the perimeter of the pipe, the effective mean stress increased greatly, from -2,860 lbs/ft 2 for a friction angle of less than 5 degrees to -27,870 lbs/ft 2 when the friction angle was set to the same value as the soil. Results are depicted graphically in Figures 6.4 and 6.5. Figure 6.4 18 Diameter FRCP Friction Area on Surface of Pipe. In situ conditions reflect a soil interacting with a concrete surface and soil-soil interaction around the pipe. In an attempt to model this, the friction analysis along the sidewall is justified but the friction around the perimeter of the pipe is questionable. It is impossible to have near zero friction around the perimeter of the pipe that would result in a stress decrease, as seen in the three-dimensional analysis for a friction angle less than 5 degrees. FRCP is a smoother pipe, when compared to SRCP, yet the friction will not be

PAGE 95

82 Figure 6.5 18 Diameter FRCP Near Zero Frictionless Area on Surface of Pipe. near zero. Thus using a friction angle equal to that of soil around the perimeter of the pipe produced more justified results. More stress is induced as friction is encountered throughout the soil depth under the load. The displacement for the interface reacted in the same manner increasing, for example, from 0.072 to 0.211 when the friction angle was increased to a value equal to that of the soil friction angle, 35 degrees. This is a direct result of the soil encountering a rough friction surface along the pipe. Table 6.5 presents the results of the three-dimensional analysis verification of the two-dimensional analysis. Note that the negative values refer to compression. Other concerns that were addressed with three-dimensional finite element analysis included stresses along the length of the pipe as a result of a possible shear stress induced on the ends of the pipe due to wall friction and displacement of the entire pipe as a result of both wall friction and service load.

PAGE 96

83 Soil Box Analysis A three-dimensional finite element analysis was conducted with and without the pipe inside the soil box to analyze the box dimensions while varying the friction angle through four different analyses. An approximate height of ten feet was used for the height of the soil box based on providing proper bedding needs for each of the 18 and 24 diameter pipe. A bedding depth of one pipe diameter below the invert of the pipe was used in the finite element analysis. Three different lengths of soil box were analyzed, ten, fifteen and twenty feet. Full-scale models were used for analyses eliminating any concerns of the results. Table 6.5 Three Dimensional Analysis Verification of Plaxis 2D Wall Friction Analysis Pipe D 1 C 2 R 3 EPS 5 EMS 6 Disp 1 IEWS 7 IEPS 8 Disp 2 PM 9 Disp 3 FRCP 18" Dense 1 -23650 -13830 0.236 -7340 -24630 -0.060 -6170 0.057 SRCP 18" Dense 1 -21600 -12690 0.237 -7280 -22440 -0.065 -5230 0.064 FRCP 18" Dense 0.05 -24430 -15400 0.331 -4840 -4590 -0.143 -961 0.090 SRCP 18" Dense 0.05 -24440 -15400 0.331 -4840 -4550 -0.143 -877 0.090 FRCP 24" Dense 1 -24520 -14690 0.230 -7410 -24960 -0.696 -11090 0.067 SRCP 24" Dense 1 -22180 -13380 0.234 -7380 -22610 -0.077 -9600 0.076 FRCP 24" Dense 0.05 -26430 -14760 0.333 -4790 -5800 -0.194 -2220 -0.113 SRCP 24" Dense 0.05 -26450 -14750 0.333 -4790 -5740 -0.195 -2070 0.113 *All values reported are extreme values (i.e., the maximum values). 1. D is the pipe diameter in inches. 2. C is the compaction of the backfill. 3. R is the interface value of strength. A value of 1 represents a friction angle the same as the backfill soil. A value of 0.05 represents a friction angle less than five degrees. 4. DM is the deformed finite element mesh after loading 16,000 lbs/ft 2 5. EPS is the effective principal stress for the entire model box in lbs/ft 2 6. EMS is the effective mean stress for the entire model box in lbs/ft 2 7. IEWS is the interface effective normal wall stress in lbs/ft 2 8. IEPS is the interface effective normal pipe stress in lbs/ft 2 9. PM is the pipe bending moment in lbs-ft/ft Disp 1 is the total displacements for the entire box in feet. Disp 2 is the total displacements for interface around the perimeter of the pipe in feet. Disp 3 is the total displacement of the concrete pipe in feet.

PAGE 97

84 The first analysis group examined the three different lengths with soil backfill only and no test pipe in the box. In this analysis, the friction angle was varied from 35 degrees to less than 5 degrees. This friction angle was assigned to the sidewalls. The remaining three analyses examined boundary effects with a 24 FRCP inserted in the soil box. Again each of the three soil box lengths was analyzed while varying the friction angle at the sidewalls and around the perimeter of the test pipe. Note that the decision to use the FRCP is not due to preference; it is merely an example to model the boundary effects since Youngs modulus for both SRCP and FRCP are similar in magnitude. Soil Box Analysis--No pipe A three-dimensional analysis was done on three different lengths of the soil box to determine the stress and displacement of the soil due to a maximum distributed load of 16,000 lbs/ft 2 An interface was assigned to the sidewalls to all variation of the friction angles. Dense compacted backfill soil was used since it provides better bedding for the test pipe than the loose compacted soil. Six models, two for each length, were analyzed with a friction value of 35 degrees and less than 5 degrees. Figures 6.6-6.8 show the stress distribution for the effective mean stresses throughout the entire soil box with sidewall friction angle 35 degrees. Figures 6.9-6.11 show the stress distribution for a sidewall friction of less than 5 degrees. For the soil box model with sidewall friction, the extreme effective mean stresses decreased as the length of the soil box increased. For the friction value of less than 5 degrees, the extreme effective mean stresses increased as the length of the soil box increased. A higher displacement resulted in higher load concentrations throughout the soil box, thus the increase in stress with an increase in length. Another observation is the stress concentration and distribution throughout the soil. In comparing the 10 soil box with

PAGE 98

85 friction to the 15 soil box with friction, the stress concentration around the middle of the box increases approximately 2,000 lbs/ft 2 This is important in dimensioning the soil box for the maximum load needed for proper testing results. Figure 6.6 10 Length Soil Box: Effective Mean Stresses with Sidewall Friction: Stresses were modeled along the sidewall. Figures 6.12-6.14 show the left sidewall deformation plane with a friction angle of 35 degrees. The sidewall interface with the largest effective normal stress occurs in the 15 soil box at a value of 7,280 lbs/ft 2 The 10 soil box has a sidewall interface stress value of 6,620 lbs/ft 2 and the 20 length a value of 7,150 lbs/ft 2 The 10 box contains only half the volume of soil as the 20 length box, resulting in the lower stress value. Also, the surface area of the 10 box exposed to the distributed load is half the area of the 20 box. Figures 6.15-6.17 show the left sidewall deformation plane with a friction angle less than 5 degree. The mean effective

PAGE 99

86 normal stress along the sidewall of the soil box with a friction angle of less than 5 degrees Figure 6.7 15 Length Soil Box: Effective Mean Stresses with Sidewall Friction. Figure 6.8 20 Length Soil Box: Effective Mean Stresses with Sidewall Friction.

PAGE 100

87 Figure 6.9 10 Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls. Figure 6.10 15 Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls.

PAGE 101

88 Figure 6.11 20 Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls. decreases as the length of the soil box increases. The stress value for the 20 soil box was 4,790 lbs/ft 2 With almost frictionless sidewalls, the soil is free to settle upon loading releasing stress from the sidewalls. In comparing the sidewall stress deformation plane for the friction wall and the frictionless wall, the former deformation plane was jagged, representing the friction on the sidewall. The frictionless walls deformation plane is smooth, representing the transition of deformation after loading.

PAGE 102

89 Figure 6.12 10 Length Effective Normal Stresses Left Sidewall Friction Plane. Figure 6.13 15 Length Effective Normal Stresses Left Sidewall Friction Plane.

PAGE 103

90 Figure 6.14 20 Length Effective Normal Stresses Left Sidewall Friction Plane. Figure 6.15 10 Length Effective Normal Stresses Left Sidewall Frictionless Plane.

PAGE 104

91 Figure 6.16 15 Length Effective Normal Stresses Left Sidewall Frictionless Plane. Figure 6.17 20 Length Effective Normal Stresses Left Sidewall Frictionless Plane

PAGE 105

92 Soil Box Analysis--Modeled with Test Pipe A three-dimensional analysis was done with a 24 concrete pipe inserted inside the box. As stated above, the depth of bedding used was one diameter of the test pipe and a dense compacted soil backfill was used. In this analysis, the area around the perimeter of the pipe and the area on the sidewalls were modeled with both a friction angle of 35 degrees and a friction angle of less than 5 degrees. Three lengths of soil box (10, 15, and 20) were analyzed using a friction angle of 35 degrees assigned around the perimeter of the pipe and along the sidewalls of the soil box. The stresses within the soil box and at the soil-structure interface were examined. The displacements of the soil and test pipe were also examined. Figure 6.18-6.20 shows the effective mean stress distribution for the entire soil box. The result of friction on the sidewalls shows a high stress concentration around the crown and invert of the pipe. In comparing the stress level between the haunch, the crown and invert of the pipe, the difference in stress values is approximately 7,000 lbs/ft 2 With a friction surface on the sidewalls, the soil is restricted in movement resulting in a more concentrated load in the center of the soil box. The sidewall maximum stress value is seen in the 20 length box. The difference in stress from the 10 to the 20 length is approximately 600 lbs/ft 2 Figure 6.21 shows the stresses for the 20 soil box, for example, imposed on the perimeter area of the test pipe. The red shading on the crown and invert of the test pipe shows a high concentration of stresses. This concentration is a result of the friction on the sidewall and the soil displacing more at the center due to the applied load.

PAGE 106

93 Figure 6.18 10 Length Effective Mean Stress with Friction on Sidewalls. Figure 6.19 15 Length Effective Mean Stress with Friction Sidewalls.

PAGE 107

94 Figure 6.20 20 Length Effective Mean Stress with Friction Sidewalls. Figure 6.21 Effective Normal Stress Imposed on Perimeter of Test Pipe. Another three-dimensional analysis was performed with a friction angle of less than 5 degrees assigned to the sidewalls and the perimeter area of the pipe. With the sidewalls having near zero friction, the soil is free to slide along the walls and also around the

PAGE 108

95 perimeter of the pipe. This analysis was done to model the stress behavior difference between the friction analyses as stated previously and the near frictionless performance of the soil. Figures 6.22-6.24show the effective mean stress concentration throughout the entire soil box due to the distributed load of 16,000 lbs/ft 2 The high stress values are concentrated around the haunches of the test pipe. Extreme effective mean stress for the 10 length is 17,320 lbs/ft 2 for the 15 length is 14,220 lbs/ft 2 and for the 20 length is 14,610 lbs/ft 2 The stress concentration decreases with increasing length. When comparing the friction and frictionless analyses, the stress concentration modeled around the perimeter of the pipe decreased by a large amount when removing the friction from around the pipe. Another observation was the doubled increase in soil displacement. This is a direct result of the soils ability to move and displace freely along the sidewalls. Total displacements for the friction surface analysis ranged from 0.2 to 0.24. The increase in displacement for the near frictionless analysis ranged from 0.4 to 0.6. A frictionless surface at the perimeter of the test pipe does not represent an accurate model for analysis. In order to examine this effect more precisely, a friction angle of less than 5 degrees was assigned to the sidewalls and a friction angle between a recommended soil/ steel structure interaction and a soil/ soil interaction was assigned to the perimeter of the pipe. An average was taken to represent a friction surface of less than the soil-soil friction yet greater than the soil-steel friction value. The results showed a more realistic performance for the structural response of the test pipe and the surrounding soil inside the soil box. Figures 6.25-6.27 show the effective mean stress distribution for this analysis.

PAGE 109

96 Figure 6.22 10 Length Effective Mean Stress with Frictionless Sidewalls/ Pipe. Figure 6.23 15 Length Effective Mean Stress with Frictionless Sidewalls/ Pipe.

PAGE 110

97 Figure 6.24 20 Length Effective Mean Stress with Frictionless Sidewalls/ Pipe. Figure 6.25 10 length Effective Mean Stress with Frictionless Sidewalls and Friction Around Perimeter of Pipe.

PAGE 111

98 Figure 6.26 15 length Effective Mean Stress with Frictionless Sidewalls and Friction Around Perimeter of Pipe. Figure 6.27 20 length Effective Mean Stress with Frictionless Sidewalls and Friction Around Perimeter of Pipe.

PAGE 112

99 The stress concentrations shown in Figures 6.25-6.27 portray a more accurate response from the soil and soil/ structure interaction. The assigned friction angles along the sidewalls and around the perimeter of the pipe effectively eliminate boundary effects. Results show no lateral stresses induce along the length of the pipe. This analysis shows a high stress concentration at the invert of the test pipe. With this high stress at the invert of the pipe, proper bedding compaction is necessary for maximum structural performance of the pipe. Four Wall Friction Analyses A three dimensional analysis examined the friction affects on the test pipe applied on all four walls of the soil box. Friction on the walls parallel to the pipe was assigned a value of 35 degrees. Though the two sidewalls were assigned friction angles, Plaxis is limited in that it does not allow friction angle specification for the front and rear walls. The test pipe was oriented perpendicular to the front and rear wall and parallel to the sidewalls. A concern was raised that if shear stress was induced on the ends of the pipe due to the test pipe sliding against the front and rear walls during loading, how would that shear stress affect the test pipe. Plaxis 3D was used to induce shear stress on the ends of the pipe and examine the effects of that stress along the length of the pipe and on the total displacement of the pipe. The friction on the front and rear wall was created using a beam element activated over a thin slice (0.01) in the z direction using the 3D mesh generation tool. Shear stress induced on the ends of the pipe became a concern for the test pipe when attempting to simulate in situ conditions. An in situ pipe will experience very little shear stress on the ends of the pipe, where-as-in a test facility, the boundary conditions pose a concern for testing of the pipe. The stresses and displacements were examined

PAGE 113

100 along the length of the pipe when the shear stress was induced on its ends as shown in Figure 6.28. Eight different analyses were performed using 18 and 24 diameter FRCP and SRCP compacted in dense soil backfill. Each diameter pipe was modeled using a sidewall friction angle of 35 degrees and less than 5 degrees. Displacements along the span length of the test pipe are shown in Table 6.7 with displacements at the front, middle, and rear of the pipe. As shown in Table 6.7, the displacements along the length of the pipe did not differ from the front to the rear. Shear stress induced at the ends of the pipe did not affect the displacements of the pipe. Figure 6.28 Plaxis 3D Test Pipe with Shear Stress Induced on Ends of Pipe The stress at the middle of the pipe due to the shear stresses encountered at the ends of the pipe posed a concern. An area around the perimeter of the test pipe modeled the effective mean stress along the full length of the pipe. The eight analyses showed very little stress increase at the mid span of the pipe. Concerns of a stress increase at the mid

PAGE 114

101 span of the pipe were analyzed using Plaxis 3D. Table 6.8 presents the extreme effective mean stress values around the mid span for each of the eight runs modeled. The stress values along the length of the pipe differed very little. To guarantee that no shear stresses are induced at the ends, the test pipe shall be 1-2 inches short of the depth of the box, allowing the pipe to settle without contact with the front or rear wall. Table 6.7 Displacement of Pipe Length with Shear Stress Induced on the Ends. Pipe Dia. R* Disp F Disp M Disp R FRCP 18" 1 0.050 0.050 0.050 SRCP 18" 1 0.050 0.050 0.050 FRCP 18" 0.05 0.019 0.019 0.019 SRCP 18" 0.05 0.019 0.019 0.019 FRCP 24" 1 0.067 0.067 0.067 SRCP 24" 1 0.067 0.067 0.067 FRCP 24" 0.05 0.035 0.035 0.035 SRCP 24" 0.05 0.029 0.029 0.029 R is the Interface strength value representing the friction angle where R=1 is a friction angle equal to the soil and R=0.05 is near zero friction. Disp F, M, R represents the pipe displacement at the front (F), Midspan (M) and Rear (R) of the pipe in feet. Table 6.8 Extreme Effective Normal Stresses Along the Length of Pipe with Shear Stress Induced on the Ends. Pipe Dia. R* EMS 1 EMS 2 FRCP 18" 1 -25750 -25620 SRCP 18" 1 -25760 -25625 FRCP 18" 0.05 -3860 -3860 SRCP 18" 0.05 -3860 -3860 FRCP 24" 1 -23220 -23540 SRCP 24" 1 -23510 -23490 FRCP 24" 0.05 -5050 -5050 SRCP 24" 0.05 -5100 -5110 R is the Interface strength value representing the friction angle where R=1 is a friction angle equal to the soil and R=0.05 is near zero friction. 1. EMS is the extreme effective mean stress on the ends of the pipe in lbs/ft 2 2. EMS is the extreme effective mean stress at the middle span of the pipe in lbs/ft 2

PAGE 115

CHAPTER 7 RECOMMENDATIONS Two and three-dimensional analyses were performed using Plaxis Finite Element Analysis for Soil and Rock modeling soil structure interaction on buried pipes. A large soil test box was designed using the stress analysis from the finite element software. Plaxis 3D provided a good examination of stress and displacement analysis on buried pipes. In this research a two-dimensional approach was initially used, which limited the analysis in the z direction. Plaxis 3D was used to analyze stress and displacements along the full length of the test pipe while buried and under loads from the cover soil and an applied distributed load. Plaxis 3D is a tunneling software program providing analysis in the z direction, the length of the test pipe. A three-dimensional approach is highly recommended for further analysis of any buried pipe research. Another concept of interest for future research is the level of compaction around the haunch of the buried pipe. Plaxis 3D will allow the user to define the compaction throughout a defined distance in the z direction running parallel to the buried test pipe. Properties for analysis would be the normal stresses, shear stresses, displacements, axial forces and bending moments along the full length of the pipe. Plaxis 3D allows the user to input the stiffness of the soil in a desired geometry formation around the pipe. A finite element analysis is recommended for the area defining a loose compaction around the haunch of the pipe for a specified distance along the test pipe. A complete analysis is recommended on the effects of poor compaction around the haunch (loose pockets). 102

PAGE 116

103 As shown in the third analysis, the highest stress concentrations occurred at the invert of the pipe assuming proper compaction. If poor compaction occurs the stress concentration locations could change, affecting the structural response of the pipe. Benefits of this analysis would include examination of the required trench width for installation of different size pipe diameters and also examination of the effect of the cover soil, whether compacted loose or dense. Other recommended analyses for Plaxis 3D are the effects of soil-structure interaction between the test pipe and surrounding soil. Again the user can specify the stiffness of the soil surrounding the test pipe for analysis of the structural performance of the test pipe. Any geometry formation can be specified in the input program for Plaxis 3D to analyze a specified area of interest. Beam elements are used for the lining of the bored tunnel opening. The 3D mesh generation is a useful tool for specific analysis along the length of the pipe. The user specifies the distance into the z-axis for analysis and the thickness of the slice. The structural performance of a test pipe can be analyzed looking at the effective normal stresses acting on the test pipe. Plaxis 3D does an excellent job of providing shaded images for the output of a finite element analysis. Stress distribution along the length of the pipe or any other material for analysis is displayed through the shading, allowing the user to visualize where the stress intensities vary in the analysis. Another good feature is the three-dimensional views in the output program. Having a three-dimensional view of a three-dimensional research project provides a better understanding of the results.

PAGE 117

104

PAGE 118

APPENDIX PLAXIS 2D ANALYSIS AND RESULTS Dimension Analysis A two-dimensional finite element analysis was conducted with and without the pipe inside the soil box to analyze the box dimensions. In the two-dimensional analysis, an approximate height of ten feet was used based on the proper bedding needs of the 24 diameter pipe. Three different widths of soil box were analyzed, ten, fifteen and twenty feet. Each analysis was first run with no pipe inside the box and then with each of the 18 and 24 diameter concrete pipes. Symmetry was used in the size analysis of the box so that only the right half of the box need be included, as shown in Figure A.1. Figure A.1 Plaxis 2D Symmetry Model of 24 diameter FRCP. Plaxis 2D modeling to determine the configuration that would minimize the sidewall stresses using the maximum load of 16,000 lbs/ft 2 The first analysis involved 105

PAGE 119

106 three different widths with no pipe in the box, just the backfill soil. Figure A.2 shows the schematic of all three box analyses, for each length, executed in Plaxis 2D. Results showed the same extreme value for the sidewall stresses for each length as 7,270 lbs/ft 2 Modeling in only two dimensions, stresses are a direct result of the load applied and the unit weight of the soil. Plaxis 2D modeled with uniform compaction resulted in a uniform stress distribution along the sidewall. In order to obtain the sidewall stresses, a cross section was taken along the sidewall displaying the extreme effective normal stress. An example is shown in figure A.3. Figure A.2 Three Different Widths Modeled in Plaxis 2D: 10, 15, 20 feet wide. A two-dimensional analysis was performed with a concrete pipe inserted in the soil box. Twenty-four inch diameter versions of both SRCP and FRCP were used to examine the stress on the sidewalls. The 24 diameter pipes were used because the stress on the sidewalls increases as the diameter of the pipe increases. Again, three different soil box

PAGE 120

107 lengths were examined, with the right half of each of the FRCP and SRCP inserted into the soil model. Stresses along the sidewalls for the ten, fifteen and twenty foot models with the FRCP are displayed in Figures A.4-A6. FRCP is displayed in the figure for illustration purposes only, both FRCP and SRCP were analyzed and resulted in similar stresses along the sidewalls. Results show that the stress decreases as box length increases, creating a maximum stress on the sidewall of 7,300 lbs/ft 2 for the twenty-foot length For verification, a full-scale finite element analysis was done for the twenty-foot length soil box model. The full-scale model reported the same stress on the sidewalls as the model using symmetry with only the right half of the SRCP and FRCP. From the 24 diameter pipe analysis, the results showed that the stresses and boundary conditions were lowest for the 20 length. Figure A.3 Example of Cross Section Used to Examine the Sidewall Stresses (20 Width).

PAGE 121

108 Wall Friction/ Soil Compaction Analysis A two-dimensional analysis examined the sidewall friction of the selected box dimension length of twenty feet, altering the friction angle of the soil structure interface with the sidewalls and the perimeter area of the pipe using two different compacted backfills. Stresses and displacements were analyzed throughout the soil and along the specified areas of interest. Two different friction angles were assigned; a friction angle of 35 degrees and a friction angle less than five degrees. I. D. Moore conducted research, as discussed in chapter 2, that indicated a minimization of the friction angle to less than five degrees helps to reduce boundary effects from inducing lateral stresses on the test pipe. The purpose is to simulate in situ conditions by minimizing the boundary condition effect on the structural response of the test pipe. A total of 24 different analyses were done to examine the wall friction and the effects of different compacted soils on the structural response of the test pipe. Three different pipe size diameters were analyzed, 18, 24, and 48, with two types of concrete pipes, FRCP and SRCP, embedded in two different compacted backfills, loose and dense. Full-scale analyses were done to eliminate any doubt within the results. The extreme value results are presented in Table A.1. It is apparent from the finite element modeling results that the dense compaction is more stable than the loose compaction. Referring to the total displacement in Table A.1, the displacement is twice as large for the loose compacted soil. Dense compacted soil will provide better bedding conditions for allowing the pipe to sustain a load. The dense compacted soil was modeled using a soil modulus value compacted to 90% proctor.

PAGE 122

109 Figure A.4 Example FRCP Cross Section of Sidewall Stresses (10 wide box). Figure A.5 Example FRCP Cross Section of Sidewall Stresses (15 wide box).

PAGE 123

110 Figure A.6 Example FRCP Cross Section of Sidewall Stresses (20 wide box). In comparing both the FRCP and SRCP, the overall difference in results was minimal. The properties examined around the pipe, such as stress and displacements in the surrounding soil, showed minimal changes between the FRCP and SRCP. When comparing the variation in friction angle, the friction angle equal to the soil friction angle resulted in an effective normal stress along the sidewall that was twice the effective normal stress of the model with a friction angle less than five degrees. It should be noted that the negative stress values represent compression.

PAGE 124

111 Table A.1 Wall Friction Analysis-Plaxis 2D Finite Element Analysis Pipe D 1 C 2 R 3 EPS 4 EMS 5 Disp 1 IEWS 6 IEPS 7 Disp 2 PM 8 Disp 3 SRCP 24" Loose 1 -24560 -14720 0.523 -7360 -24990 -0.157 -10550 0.151 FRCP 24" Loose 1 -24550 -14710 0.523 -7450 -24990 -0.157 -11100 0.152 SRCP 24" Dense 1 -24550 -14710 0.231 -7380 -24980 -0.069 -10550 0.067 FRCP 24" Dense 1 -24530 -14700 0.230 -7260 -24970 -0.070 -10950 0.067 SRCP 24" Loose 0.05 -26430 -14760 0.755 -4790 -5800 -0.442 -2090 0.257 FRCP 24" Loose 0.05 -26430 -14760 0.755 -4790 -5800 -0.442 -2190 0.257 SRCP 24" Dense 0.05 -26430 -14760 0.333 -4790 -5800 -0.195 -2090 0.113 FRCP 24" Dense 0.05 -26430 -14760 0.333 -4790 -5800 -0.194 -2190 0.113 SRCP 18" Loose 1 -23690 -13850 0.535 -7340 -24660 -0.135 -5780 0.129 FRCP 18" Loose 1 -23680 -13840 0.535 -7340 -24660 -0.135 -6070 0.129 SRCP 18" Dense 1 -23680 -13840 0.236 -7340 -24650 -0.059 -5770 0.057 FRCP 18" Dense 1 -23660 -13840 0.236 -7340 -24640 -0.059 -6070 0.057 SRCP 18" Loose 0.05 -24430 -15400 0.750 -4840 -4590 -0.325 -884 0.204 FRCP 18" Loose 0.05 -24430 -15400 0.750 -4840 -4590 -0.324 -961 0.204 SRCP 18" Dense 0.05 -24430 -15400 0.331 -4840 -4590 -0.143 -884 0.090 FRCP 18" Dense 0.05 -24430 -15400 0.331 -4840 -4590 -0.143 -941 0.090 SRCP 48" Loose 1 -23300 -13840 0.483 -7120 -22920 -0.289 -40150 0.283 FRCP 48" Loose 1 -23290 -13840 0.482 -7120 -22920 -0.289 -41560 0.282 SRCP 48" Dense 1 -23290 -13840 0.213 -7120 -22910 -0.127 -40140 0.125 FRCP 48" Dense 1 -23280 -13840 0.213 -7120 -22910 -0.127 -41550 0.125 SRCP 48" Loose 0.05 -30550 -17710 1.230 -5130 -11990 -1.130 -18600 0.624 FRCP 48" Loose 0.05 -30550 -17710 1.230 -5130 -11990 -1.130 -19370 0.624 SRCP 48" Dense 0.05 -30550 -17710 0.541 -5130 -11990 -0.499 -18600 0.275 FRCP 48" Dense 0.05 -30550 -17710 0.540 -5130 -11990 -0.499 -19240 0.275 **All values reported are extreme values (i.e., the maximum values). 1. D is the pipe diameter in inches. 2. C is the compaction of the backfill. 3. R is the interface value of strength. A value of 1 represents a friction angle the same as the backfill soil. A value of 0.05 represents a friction angle less than five degrees. 4. EPS is the effective principal stress for the entire model box in lbs/ft 2 5. EMS is the effective mean stress for the entire model box in lbs/ft 2 6. IEWS is the interface effective normal wall stress in lbs/ft 2 7. IEPS is the interface effective normal pipe stress in lbs/ft 2 8. PM is the pipe bending moment in lbs-ft/ft Disp 1 is the total displacements for the entire box in feet. Disp 2 is the total displacements for interface around the perimeter of the pipe in feet. Disp 3 is the total displacement of the concrete pipe in feet.

PAGE 125

LIST OF REFERENCES American Society for Testing and Materials, Acoustic Emission. ASTM E 1932-97, Philadelphia, 2001. Beattie Alan G, Experimental methods-nondestructive evaluation techniques, Handbook on Structural Testing, Society for Experimental Mechanics, Inc., The Fairmont Press, Inc., Lilburn, GA, pp. 237-294, 1993. Bishop, R R, Time dependent performance of buried PVC pipe. Underground Plastic Pipes, International Conference of Underground Plastic Pipe, March 30-April 1, 1981, New Orleans, Louisiana, pp. 202-212. Bland, C E G, Sheppard, K J, Investigations into the structural performance of clay pipes, Advances In Underground Pipeline Engineering, Proceedings of the International Conference, Jeyapalan, J K (ed), University of Wisconsin-Madison, 1985, pp. 100-116. Brinkgreve R B J, Plaxis Manual. Plaxis, Rotterdam, Netherlands, 1998. Brachman et al. R W I, Moore I D, Rowe R K, 2000, The design of a laboratory facility for evaluating the structural response of small diameter buried pipes, Canadian Geotechnical Journal, 37: 281-295. Brachman et al. R W I, Moore I D, Rowe R K, 2001, The performance of a laboratory facility for evaluating the structural response of small diameter buried pipes, Canadian Geotechnical Journal, 37: 260-275. Brachman et al. R W I, Moore I D, Rowe R K, 1996, Interpretation of a buried pipe test: small diameter pipe in the Ohio University facility, Transportation Research Record 1541, Structures, Culverts, and Tunnels. pp. 64-70. Fintel Mark 1974, Handbook of Concrete Engineering. Van Nostrand Reinhold Company, New York, NY. Gaube, E, Mueller, W, years of deformation measurement on plastic sewer pipes from Hostalen GM 5010, Underground Plastic Pipe, International Conference on Underground Plastic Pipe, March 30-April 1, 1981, New Orleans, Louisiana, pp. 288-297. 112

PAGE 126

113 Minolta USA. Minolta 3D Digitizer. Minolta Co. Ltd. January 9, 2003 Molin, J, Flexible pipes in buried clay, Underground Plastic Pipe, International Conference on Underground Plastic Pipe, March 30-April 1, 1981, New Orleans, Louisiana, pp. 322-337. Moser A P, 2001, Buried Pipe Design. 2 nd Edition McGraw Hill, New York, NY, pp. 65-82. Parmalee, R A, A study of soil structure interaction of buried concrete pipe, Concrete Pipe and the Soil-Structure System, ASTM STP 630, Bealey, M, Lemons, J D, editors, American Society for Testing and Materials, 1977, 99. 66-75. Scott Ian G, 1991, Basic Acoustic Emission. Gordon and Breach Science Publishers, New York, pp. 43-83. Selander, C E, Hickey, M E, Causey, F E, Howard, A K, Evaluation of reinforced plastic mortar pipe a government industry cooperative study, REC-ERC-72-26, Bureau of Reclamation, United States Department of the Interior, 1972. Singhal, A C, Veliz, V, Experimental and field observation of dynamic behavior of buried pipelines, Advances in Underground Pipeline Engineering, Proceedings of the International Conference, Jeyapalan, J K (ed), University of Wisconsin-Madison, 1985, pp. 302-310. Todres, H A, McClinton, M, Stress and strain responses of a soil-pipe system to vehicular traffic, Advances In Underground Pipeline Engineering, Proceedings of the International Conference, Jeyapalan, J K (ed), University of Wisconsin-Madison, 1985, pp. 428-437. Vary, Alex, 1988, The acousto-ultrasonic approach, Acousto-Ultrasonics Theory and Application. Plenum Press, New York and London.

PAGE 127

BIOGRAPHICAL SKETCH Melissa Kay Crosby was born April 30, 1978, in Naples, Florida, to Lacy Oler and Sandra Kay Crosby. She graduated from Santa Fe High School in May 1996 in the top 5% of her class and started her college career at Santa Fe Community College in Gainesville, Florida, on a full paid scholarship. She received her Associate of Arts degree in August 1999 and transferred to the University of Florida to pursue a Bachelor of Science in Civil Engineering in the summer of 1999. While attending the University of Florida full time, Melissa worked part time for a local civil engineering firm, Kelley Engineering, Inc., for two years interning and gaining valuable experience in the work field. She received her Bachelor of Science in Civil Engineering in December 2001, graduating with honors. Melissa continued with her education entering graduate school to pursue a Master of Engineering in the Materials Group of the Civil and Coastal Engineering Department in January 2002 immediately following her undergraduate completion. After graduating from the University of Florida with a Master of Engineering, Melissa plans on moving to Atlanta, Georgia, to pursue a career in the exciting and challenging field of civil engineering design. Melissa also plans to become licensed as a professional engineer in the state of Georgia. 114


Permanent Link: http://ufdc.ufl.edu/UFE0000669/00001

Material Information

Title: Finite element analysis of a laboratory soil box test facility for evaluating the structural response of concrete pipe
Physical Description: Mixed Material
Creator: Crosby, Melissa Kay ( Author, Primary )
Publication Date: 2003
Copyright Date: 2003

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000669:00001

Permanent Link: http://ufdc.ufl.edu/UFE0000669/00001

Material Information

Title: Finite element analysis of a laboratory soil box test facility for evaluating the structural response of concrete pipe
Physical Description: Mixed Material
Creator: Crosby, Melissa Kay ( Author, Primary )
Publication Date: 2003
Copyright Date: 2003

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0000669:00001


This item has the following downloads:


Full Text












FINITE ELEMENT ANALYSIS OF A LABORATORY SOIL BOX TEST FACILITY
FOR EVALUATING THE STRUCTURAL RESPONSE OF CONCRETE PIPE















By

MELISSA KAY CROSBY


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Melissa Kay Crosby

































This thesis is dedicated to my loving family, my parents, Sandra and Oler Crosby, my
sister Stacy Crosby and to my loving fiance Devin Drake. I would also like to dedicate
this thesis to my loving Aunt Linda and Uncle Dewayne as they have offered their
support and love throughout this endeavor. It is with the love and support of my family
and friends that I am able to reach my goals.















ACKNOWLEDGMENTS

I would like to thank all of the members of my supervisory committee for their help

and ideas throughout this effort. Dr. David Bloomquist, committee cochair, provided

much insight, knowledge and financial support toward the completion of the work. Dr.

Andrew Boyd, committee chair, provided valuable time and knowledge of the subject, as

well as financial support, making this research successful.

I would also like to thank Dr. H.R. Hamilton for his contribution of time and

knowledge, which provided to be invaluable assistance during this effort. An additional

thank you goes to Scott Jacobs for his time and the computer expertise he provided

during the research.

Above all, I would like to thank God for giving me the ability to withstand the trials

and tribulations throughout this effort. It is through Him that I am able to persevere and

succeed in all my endeavors.

I would also like to thank my best friend, who is also my fiance, Devin Drake, for

the support and patience he offered me during the research and writing of this thesis.
















TABLE OF CONTENTS
page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ............................ .............................. .. .... ........ .... vii

LIST OF FIGURES ................................................. ........ ........... viii

ABSTRACT .............. ..................... .......... .............. xii

CHAPTER

1 IN TRODU CTION ................................................. ...... .................

2 LITER A TU R E REV IEW ............................................................. ....................... 3

Introduction to Pipes ............................................................ .. ........ ..
C o n c rete P ip e s .............................................................................................................. 3
Concrete Pipe Testing ............................................. .............. ..... .......... 4
Three-E dge B hearing Strength ........................................................................... 6
B wedding Factors and Classifications................................... ............................. ....... 7
H history of Pipe Testing Facilities ................................... ........................... ................9

3 STANDARD REINFORCED CONCRETE PIPE VS. FIBER REINFORCED
CON CRETE PIPE ....................................................... .............. .. 22

Background Information on Concrete Pipes................................... ...............22
Cracking in Concrete--The Fracture Zone.................. .... ........... .............. 23
Standard Reinforced Concrete Pipe--SRCP .................................... ............... 25
M echanics of SR C P ........2......_.. ......... .. .. ................ .. .............. 25
R e in fo rc em en t ............................................................................................... 2 7
Strength..................................................27
Structural P perform ance ......................................................................... ....... ........28
Fiber Reinforced Concrete Pipe--FRCP .......................................... ............... 29
Mechanics of FRCP...................... ..............................30
Fiber-M atrix Bond .................. .......................... .... .... ................3 1
Fiber-Fiber Interaction ................ .... ................................... .. .. ........ ..... 31
Load vs. Deflection in Fiber Reinforced Concrete ...........................................32
Stages of Cracking-Fiber Intervention ..................................... ............... 33
S tre n g th ............................. ......................................................... ............... 3 5
T o u g h n e ss ....................................................................................3 5


v









FR CP versus SR CP.................. ................................ ...... .. ........ .... 36
M manufacturing ............................................ .. .. ............. ......... 36
In stallation ..................................................................................................36
P e rfo rm a n c e ................................................................................................... 3 7

4 INSTRUMENTATION AND TESTING TECHNIQUES .........................................38

Crack Detection and Deflection in Concrete Pipes ............... ..................... .........38
Background on Acoustic Emission Testing..................................... ............... 38
Pream plifiers................................... .............................. .......... 42
Postam plifiers and Signal Processors............................................................... 42
Transient Recorders .................. .......................... ...... ................... 43
Spectrum A nalyzers........... ......................................................... .. .... ..... ....44
LAM --Local A rea M monitor ............................................................ ............... 45
M inolta 3D D igitizer--VIV ID 900......................................... ......................... 48
H ardw are ....................................................... 4 8
Accessories ................ ......... ................... 49
Operation ......... ......... ......... .. ...............51

5 PLAXIS VERSION 7.2--FINITE ELEMENT CODE FOR SOIL AND ROCK
A N A L Y S IS ........................................................................................................... 5 3

Introduction .......................................................................................................53
P lax is--In p u t ................................................................5 6
Plaxis--C calculations ....................... ......... ........ .. .............................64
P lax is--O u tp u t ................................................................................................ 6 8

6 FINITE ELEMENT ANALYSIS OF A SOIL BOX TEST FACILITY ....................74

In tro du ctio n .................. ........................................................................................ 7 4
Plaxis 3D --V erification Analysis................................ ................... 77
S o il B o x A n aly sis ................................................................................................. 8 3
Soil Box Analysis--No pipe............................................ 84
Soil Box Analysis--Modeled with Test Pipe ..................................92
Four W all Friction A nalyses ....................................................... 99

7 R E C O M M E N D A TIO N S .................................................................................... 102

APPENDIX PLAXIS 2D ANALYSIS AND RESULTS ..............................................105

D im en sion A n aly sis ..................................................................................... 10 5
W all Friction/! Soil Compaction Analysis................................ ........... ....108

LIST OF REFERENCES .................. ........ .................112

BIOGRAPHICAL SKETCH ............... ......... ........ ........114
















LIST OF TABLES


Table page

3.1 Typical Fiber-Matrix Pullout Strengths (Mindess, 2003). .....................................32

4.1 Specifications of Minolta VIVID 900 3D digitizer ......... .................................. 52

6.1 M material Properties of the soil (Loose & Dense).................. ............ ....................76

6.2 Material Properties of the 18" diameter concrete pipes (FRCP & SRCP) ..............76

6.3 Material Properties of the 24" diameter concrete pipes (FRCP & SRCP) ..............76

6.4 Material Properties of the 48" diameter concrete pipes (FRCP & SRCP) ..............77

6.5 Three Dimensional Analysis Verification ofPlaxis 2D Wall Friction Analysis .....83

6.7 Displacement of Pipe Length with Shear Stress Induced on the Ends.................101

6.8 Extreme Effective Normal Stresses Along the Length of Pipe with Shear Stress
Induced on the Ends. .............................................. .... .. ...... ........101

A. 1 Wall Friction Analysis-Plaxis 2D Finite Element Analysis...............................111
















LIST OF FIGURES


Figure page

2.1 E arly C concrete Pipe Testing............................................................ ............... 5

2.2 Three edge bearing test for concrete pipe ........................................ .....................7

2.3 O hio U university Full Scale Testing Site................................................................ 13

2.4 The Center for Pipes and Underground Structures Test Facility at Ohio University. 13

2.5 Hardie Pipe's Rigid Soil Box Front View...........................................................15

2.6 Hardie Pipe's Rigid Soil Box Full Image View..................... ............. ............... 16

2 .7 H ardie P ipe F flexible Soil .............................................................. ..................... 16

2.8 Hardie Pipe Flexible Soil Box Testing conducted at UCF ..................................... 17

2.9 LVDT's Shown to Measure Deflection of Sidewalls ..............................................18

2.10 Flexure Crack At The Crown of Fiber Reinforced Concrete Pipe.......................... 19

2.11 Reinforced Concrete Pipe Cracking At The Crown...............................................20

3.1 Coordinate system and stress components ahead of crack tip..............................24

3.2 Three M odes of Cracking ...................................... ......... ................... 24

3.3 Standard Reinforced Concrete Pipe Section Cracked...................... ...............26

3.4 Standard Reinforcing Steel Rebar. ........................................ ........................ 28

3.5 Typical Fiber Reinforced Concrete Pipe..........................................................29

3.6 Typical load-deflection curve for fiber reinforced concrete in flexure .................33

3.7 Schematic representation of fibers bridging across a crack. ...................................34

4.1 Burst Acoustic emission signal with properties ....................................... .......... 39

4.2 A acoustic Em mission Process............................................. .............................. 40









4.3 Pre-am plifiers ............................................................. .... .......... 43

4.4 Transient recorder with multiple AE signals................................. ...... ............ ...44

4.5 D digital O scilloscope ............................................ ................. .. ...... 45

4.6 LAM--Local Area Monitor .......................................46

4.7 Minolta VIVID 900 Non-Contact 3-D Digitizer.....................................................49

4.8 Com pact Flash M em ory Card ............................................................................ 49

4.9 Rotating Stage Set for Scanning a Full 3-D image ..............................................50

4.10 Tripod (left) and Tilting Base Mount (right) for Minolta VIVID 900...................50

5.1 Plaxis 7.2 Computer Aided Drafting screen used to create modeling analysis........54

5.2 General Settings window in Plaxis 7.2.......................................... ............... 56

5.3 Plaxis 7.2 M ain Toolbar. ............................................... ................................ 58

5.4 Tunnel Designer in Plaxis 7.2. .................................. ..................................58

5.5 Standard Fixities in Plaxis 7.2 shown on a soil box with right half tunnel .............59

5.6 M material Sets W indow in Plaxis 7.2 ....................................................................... 61

5.7 Soil Input in Plaxis 7.2--Mohr Coulomb Model. .............................................. 61

5.8 Beam Properties Input Window in Plaxis 7.2. ................................. .................62

5.9 Pore Water Pressure & Initial Stress Modes in Plaxis 7.2. .....................................63

5.10 Plaxis 7.2 Calculations Program ........................................ ......................... 65

5.11 Plaxis 7.2 Calculations Program--Parameters Tab.............................................66

5.12 Plaxis Calculations Program--Multipliers Tab........ .........................67

5.13 Plaxis Output program with deformed mesh displayed on example model.............69

5.14 Plaxis Output Effective Mean Stresses Displayed by Mean Shading....................69

5.15 Plaxis Output Effective Mean Stresses Displayed by Contours. ..........................70

5.16 Stress Distribution Cross Section A-A in Plaxis 7.2--Output Program ...................71

5.17 Horizontal Displacement Cross Section A-A in Plaxis 7.2--Output Program .........72









5.18 Displacements for the Pipe and Interface--Plaxis 7.2 Output Program. .................72

5.19 Bending Moment for the Pipe--Plaxis 7.2 Output Program...............................73

6.1 Three Dimensional View of Total Displacements for 24" FRCP..........................79

6.2 Three Dimensional View of Total Displacements for 24" SRCP............................79

6.3 Left Side Interface of Soil Box M odel 24" FRCP .......................................... 80

6.4 18" Diameter FRCP Friction Area on Surface of Pipe. ................. .................81

6.5 18" Diameter FRCP Near Zero Frictionless Area on Surface of Pipe .....................82

6.6 10' Length Soil Box: Effective Mean Stresses with Sidewall Friction: ..................85

6.7 15' Length Soil Box: Effective Mean Stresses with Sidewall Friction....................86

6.8 20' Length Soil Box: Effective Mean Stresses with Sidewall Friction....................86

6.9 10' Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls............87

6.10 15' Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls............87

6.11 20' Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls............88

6.12 10' Length Effective Normal Stresses Left Sidewall Friction Plane .....................89

6.13 15' Length Effective Normal Stresses Left Sidewall Friction Plane .....................89

6.14 20' Length Effective Normal Stresses Left Sidewall Friction Plane .....................90

6.15 10' Length Effective Normal Stresses Left Sidewall Frictionless Plane. ................90

6.16 15' Length Effective Normal Stresses Left Sidewall Frictionless Plane ...............91

6.17 20' Length Effective Normal Stresses Left Sidewall Frictionless Plane ..............91

6.18 10' Length Effective Mean Stress with Friction on Sidewalls.............................93

6.19 15' Length Effective Mean Stress with Friction Sidewalls.............................. 93

6.20 20' Length Effective Mean Stress with Friction Sidewalls.............................. 94

6.21 Effective Normal Stress Imposed on Perimeter of Test Pipe.................................94

6.22 10' Length Effective Mean Stress with Frictionless Sidewalls/ Pipe ....................96

6.23 15' Length Effective Mean Stress with Frictionless Sidewalls/ Pipe. ...................96









6.24 20' Length Effective Mean Stress with Frictionless Sidewalls/ Pipe ....................97

6.25 10' length Effective Mean Stress with Frictionless Sidewalls and Friction Around
Perim eter of Pipe ......... .... .................................. ..... ...... .. ........ .... 97

6.26 15' length Effective Mean Stress with Frictionless Sidewalls and Friction Around
Perimeter of Pipe ............. ...... .................. ........ .... ......... 98

6.27 20' length Effective Mean Stress with Frictionless Sidewalls and Friction Around
Perimeter of Pipe ............. ...... .................. ........ .... ......... 98

6.28 Plaxis 3D Test Pipe with Shear Stress Induced on Ends of Pipe ........................... 100

A.1 Plaxis 2D Symmetry Model of 24" diameter FRCP. ......................... ...........105

A.2 Three Different Widths Modeled in Plaxis 2D: 10, 15, 20 feet wide................. 106

A.3 Example of Cross Section Used to Examine the Sidewall Stresses (20' Width)...107

A.4 Example FRCP Cross Section of Sidewall Stresses (10' wide box)....................109

A.5 Example FRCP Cross Section of Sidewall Stresses (15' wide box)....................109

A.6 Example FRCP Cross Section of Sidewall Stresses (20' wide box)....................110















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

FINITE ELEMENT ANALYSIS OF A LABORATORY SOIL BOX TEST FACILITY
FOR EVALUATING THE STRUCTURAL RESPONSE OF CONCRETE PIPE

By

Melissa Kay Crosby

May 2003

Chair: Andrew J. Boyd
Cochair: David Bloomquist
Major Department: Civil and Coastal Engineering

Two and three dimensional finite element analyses were used to examine the

stresses on a laboratory soil box test facility designed to evaluate the structural response

of fiber reinforced and standard reinforced concrete pipes. Boundary conditions were

examined in an attempt to minimize the effects on the concrete pipe tested. The state of

stress in the soil due to the applied loading was studied, along with its influence on the

concrete pipe. An extensive literature search is presented on the history of soil box

testing of pipes. Small and large-scale test facilities have been discussed through

previous testing in the literature search. A description of the test facilities, both small and

large, is presented. Plaxis, a two and three-dimensional finite element analysis program,

analyzes the stresses on the sidewalls for the dimension design of a soil box properly

scaled to minimize the boundary effects on the concrete pipe specimens. Comparisons

were made among three different proposed box lengths. It was found that as the length









increased, the stress concentrations and intensities decreased, thus minimizing the

boundary effects.

Sidewall friction and its effects on the test pipe were examined. Two and three

dimensional finite element analysis was used to assign a friction angle on the sidewalls in

order to model the soil structure interaction. An additional analysis was done to examine

the possibility of shear stresses induced on the ends of the pipe. Friction was induced on

the front and rear wall to create shear stresses on the ends of the pipe. Shear stresses

were induced on the ends of the pipe in the three-dimensional analyses but very little

difference in displacement occurred. A comparison of soil backfill, between a loose and

dense compaction, showed a large difference in overall settlement. Maximum stresses

and pipe deflections were approximately double for the loose compaction when

compared to the dense backfill.














CHAPTER 1
INTRODUCTION

Structural performance testing of small diameter buried pipes dates back to the

1930's. Large scale testing of buried concrete pipes provides useful information for

evaluating the soil structure response expected under field conditions. In the United

States, large-scale test facilities exist at Utah State University, the University of

Massachusetts at Amherst, and Ohio University. Additional facilities can be found at the

University of Western Ontario in Canada and LGA Geotechnical Institute in Germany

(Brachman et al., 2001). Two small-scale test facilities have been designed and

constructed by Hardie Pipe, Inc. and are pending patents. Each of the testing facilities at

these institutions suffers from limitations stemming from boundary conditions inherent in

their respective equipment configurations. Many of the existing test facilities cannot

closely approximate expected field conditions with respect to the stress states associated

with small diameter buried pipes. When large scale testing is not possible at a particular

location, a laboratory test facility is needed to examine the structural response of buried

pipes. A soil box test facility is required to investigate small diameter buried pipes and

allow a laboratory assessment of their performance under simulated field conditions.

Laboratory test facilities allow better control over monitoring and testing

procedures than do field testing techniques. Boundary conditions, such as the method of

loading and the geometry of the test facility, may significantly influence the results of the

test. The structural response of buried pipe can be significantly affected by the boundary

conditions if the test facility is not properly designed. Sufficient dimensions are









necessary, under expected loading conditions, to limit sidewall deflection and minimize

boundary condition effects. The primary concern to be considered in the analysis of

testing apparatus design is elimination of the influence of boundary conditions on the

structural response of the concrete pipe specimen.

Both a two and three-dimensional finite element analysis program (i.e., Plaxis 7.2

and Plaxis 3D Tunnel) will be used to evaluate the boundary conditions induced by an

applied distributed load. Stresses imparted by the load dissipate throughout the soil,

applying load to the pipe, which is surrounded by backfill. A wall stress and sidewall

friction analysis will provide dimensional information that can be used to minimize the

effect of boundary conditions. Two different backfills will be evaluated, loose and

densely compacted soil.

The objective of this work is to perform a finite element analysis for a laboratory

test facility to be used in the evaluation of structural performance in small diameter

concrete pipes, including both fiber reinforced and traditional steel reinforced concrete.

The facility will be fabricated of steel and designed to limit sidewall deflections. The soil

box test facility will be loaded using inflatable bladders placed on top of the soil.

Attention will be focused on the influence of the boundary conditions on the soil box and

how efficiently and accurately the test setup reproduces actual field conditions for small

diameter buried pipe. Consideration will be given to determination of optimal test cell

dimensions and, the influence of sidewall friction and boundary stiffness on the structural

performance of the concrete pipe and soil behavior.














CHAPTER 2
LITERATURE REVIEW

Introduction to Pipes

Buried pipes and/or conduits have improved the standard of living for people since

the beginning of civilization. Remnants of such structures from ancient civilizations have

been found in Europe, Asia, and even the western hemisphere, where some of the ancient

inhabitants of South and Central America established functional water and sewer systems

(Moser, 2001). Buried pipes serve many purposes, including sewer lines, drain lines,

water mains, gas lines, telephone and electrical conduits, culverts, oil lines, coal slurry

lines, subway tunnels, and heat distribution lines. In comparing the design used in the

1800's to the design applications we have today, it is apparent that the degree of

technology has increased significantly.

Engineers and planners take subsurface infrastructure into account before

developing buildings and houses for a community. The underground water systems serve

as arteries for cities, and the sewer systems serve as veins to carry off the waste (Moser,

2001). High quality drinking water is taken for granted by humans in today's society. To

ensure adequate quality, pipes must be designed and constructed to prevent the

introduction of contaminants. The same standards apply to sewer pipes so as to prevent

seepage of contaminants into the ground, which may reach the water table and aquifers.

Concrete Pipes

Pipes are classified as either rigid or flexible. A flexible pipe is defined as one that

will deflect at least 2 percent without structural distress. Flexible pipes, such as those









made of polyethylene plastics, do not fall within the scope of this project. A rigid pipe is

one that does not meet the flexible pipe deflection criterion. The two major types of rigid

concrete pipes are steel reinforced concrete pipes and fiber reinforced concrete pipes.

Parameters of the pipes are analyzed to design for maximum performance. Rigid pipes

must have the strength to resist wall stresses due to internal pressures or external loads

that are considered critical under anticipated service conditions. Design parameters

include strength, stiffness, corrosion resistance, density, durability and ease of joining. A

pipe must have sufficient strength and/or stiffness to perform its intended function and

also must be sufficiently durable to perform this function throughout its intended service

life. Strength is defined as the ability to resist applied stress. Internal pressures, soil

pressures, live loads, differential settlement and longitudinal bending moments impose

stress on the pipe. Stiffness is described as a material's ability to resist deflection. The

modulus of elasticity of a material is directly related to its stiffness, thus affecting

deformation of the pipe wall during loading. Durability is the ability of a material to

resist degradation (i.e., corrosion, chemical deterioration, abrasion) due to deleterious

environmental exposures. Durability is a critical parameter when determining a design

service life for performance.

Concrete Pipe Testing

In the early 1900's, concrete pipe testing consisted of placing sand bags on top of a

pipe to obtain a static distributed load as shown in Figure 2.1. In 1913, Marston

researched loads induced on buried pipe by the soil above the pipe. This research marked

the beginning of the development of a method for calculating earth loads on buried pipes.

Shortcomings include an assumption that the vertical load due to the backfill being

























Figure 2.1: Early Concrete Pipe Testing (Photo courtesy of Hardie Pipe, Inc.)

uniformly distributed. Results were significantly limited by the technology available at

the time and did not include such effects as pipe- soil interaction or settlement ratio. The

Marston load theory, based upon the concept of a prism-shaped soil load being applied to

the pipe, resulted in what is known today as the Marston load equation (2.1).

Wd = CdB2 (2.1)

Wd= load on conduits per unit length

Cd = load coefficient for ditch conduits

7= unit weight of backfill

Bd = horizontal width of ditch at top of conduit

As technology progressed, more test methods were standardized for the evaluation

of concrete pipe strength. Spangler conducted research in the 1930's that proposed four

classifications of bedding for pipes covering normal installations in the field. In order to

investigate the soil-structure interaction, the American Concrete Pipe Association

(ACPA) undertook a long-range research program to examine the nature of the loading

imposed on a buried pipe (Moser, 2001). Their research covered the development of a









finite element program used to simulate non-linear behavior of buried pipe, validation of

the program, and analysis of the soil around the pipe. Full scale testing conducted at the

Transportation Research Center in East Liberty, Ohio found the strains along the length

of the pipe to be insignificant. The results were symmetric about the vertical plane of

symmetry of the pipe, thus validating the use of plane strain finite element analysis.

Three-Edge Bearing Strength

Rigid nonpressure pipes are tested for strength in the laboratory using the three-

edge bearing test (ASTM C 497). The performance criteria require the each pipe size and

class pipe to reach specified laboratory strengths relative to the anticipated service load

condition and required ultimate strength. Traditional design practice uses the three-edge

bearing load that produces a 0.01-inch crack width as the design load. The failure load in

three-edge bearing test is defined as the load per unit length required to cause crushing or

critical cracking of the pipe test specimen. The strength thus obtained is the failure load

in the laboratory only, and is not necessarily equivalent to the load that will cause failure

in the field under buried conditions. Figure 2.2 shows a schematic diagram of the three

edge bearing test for a rigid pipe, where W represents the distributed load, D the pipe

diameter, R the radius of the wood support blocks and C the clearance beneath the pipe.

Testing of nonreinforced concrete pipes are specified in ASTM C 14. Nonpressure

reinforced concrete pipe is specified by its "D-load" strength, as determined in ASTM C

76. The D- load is defined as the load applied to a pipe under three-edge bearing

conditions, expressed in pounds per linear foot per foot of inside diameter.














SJy













Figure 2.2: Three edge bearing test for concrete pipe

Bedding Factors and Classifications

Laboratory testing and field-testing can result in two different strengths for

concrete pipes. As previously stated, the strength causing failure in the laboratory does

not always cause failure in a buried condition. Past experiments show that the estimated

load required to cause failure of a buried pipe, according to the Marston load equation, is

greater than that resulting from the three-edge bearing strength.

The most important factor influencing this discrepancy is the method in which the

pipe is bedded. Bedding factors are variables that were developed to account for the type

of soil in which the pipe is installed. The bedding factor (sometimes called the load

factor) is the ratio between the strength of a buried pipe and the strength of the same pipe

as determined by the three-edge bearing test. Bedding conditions affect the support

reaction beneath the pipe and the lateral pressure on the pipe.









Conduits used in ditch drainage have four bedding classifications. The load factors

associated with these classifications have been determined empirically and do not take

into account any lateral pressures exerted by the backfill. Furthermore, it has been noted

that the specified soil compaction cannot be depended upon reliably. Bedding

classifications in which a licensed engineer should inspect the installation are Class A,

Class B, Class C and Class D.

Class A (also known as Concrete Cradle Bedding has a load factor of 2-4 and

occurs when the lower part of the conduit is bedded in a cradle constructed of 2,000 psi

concrete, having a minimum thickness of one-fourth the pipe's internal diameter. The

cradle must extend up the sides of the pipe for a height equal to one-fourth its outside

diameter.

Class B (also known as First Class Bedding) has a load factor of 1.9 and is where

the pipe is carefully bedded in an earth foundation, composed of fine granular materials,

that is carefully shaped to fit the lower part of the pipe. The minimum bedding width is

60 percent of the pipe diameter and the remainder of the conduit must be entirely

surrounded to a height at least 1 ft above its top by granular materials that are carefully

placed in order to completely fill all spaces under and adjacent to the pipe. This fill

material must be thoroughly compacted on each side, and beneath the pipe, and placed in

layers not exceeding 0.5 ft in thickness.

Class C (also known as Ordinary Bedding) has a load factor of 1.5, and applies to

pipe bedded with "ordinary" care in an earth foundation shaped to fit the lower part of the

pipe, with reasonable accuracy, for a width of at least 50 percent of its outside diameter.

The remainder of the pipe is surrounded to a height of at least 0.5 ft above its top by









granular materials that are shovel-placed and shovel-tamped to completely fill all spaces

under and adjacent to the pipe (Moser, 2001).

Class D (also known as Impermissible Bedding) has a load factor of 1.1. This class

is where there is little or no care applied to shaping the foundation to fit the lower part of

the pipe or to refill all spaces under and around it. This class of bedding is not

recommended for culvert and sewage pipe. Major pipe manufacturing associations

recommend bedding factors that correspond to those listed in the Water Pollution Control

Federation Manual of Practice, No. FD-5, Gravity Sanitary Sewer Design and

Construction (Moser, 2001).

History of Pipe Testing Facilities

A pipe's insitu performance is a function of its material properties, any applied

loads, and the soil-structure interaction. Concrete pipe testing dates back to the early

1900's, when testing involved the placement of sand bags on top of the pipe to achieve

static, distributed load. Traditional concrete design methods are based on research

conducted at Iowa State University that dates back to the 1930's. By the 1930's,

Spangler had proposed four classifications of bedding that covered the range of

installations that could be anticipated in normal installations (Parmalee, 1977). Some

design methods were found to be too conservative as research progressed.

Prior to the 1970's, the testing of buried pipes, along with the soil-structure

interaction, did not produce accurate results. In 1970, the American Concrete Pipe

Association undertook a long-range research program to determine the nature of loading

imposed on buried concrete pipe and to develop a reliable design method based upon

soil-structure interaction (Parmalee, 1977). This research program purported to develop a

comprehensive finite element analysis program to simulate the non-linear behavior of









buried concrete pipe. The results of the research program indicated that strains along the

length of the pipe were insignificant. It was also determined that results were symmetric

about the vertical plane of symmetry of the pipe, thus validating the use of plane strain

finite element analysis.

Soil box testing began in the 1960's. This approach consists of a three-dimensional

box of known size, containing a pipe and backfilled with soil. Between 1960 and 2000,

several researchers conducted large and small scale testing on pipes under simulated

insitu conditions. Each of these research programs were formulated so as to design a test

that would yield results similar to those found in the field.

Selander et al. evaluated reinforced plastic mortar pipe in 1972. His design for the

height of overlying soil (approximately 12 feet high) proved later to be overly

conservative.

Bland and Sheppard (Bland, 1985) used research findings from the Transport and

Road Research Laboratory and the Clay Pipe Development Association obtained with

unrealistic boundary conditions to develop a test for clay pipes using a large test pit in

order to investigate structural performance. By using a large test pit, the prior limitations

of unrealistic boundary conditions found in a small soil box test were eliminated.

Bishop tested buried PVC pipes in a soil cell and in a full-scale embankment. He

proved it is possible to separate the long-term behavior of the pipe from the long-term

behavior of the soil. In-situ load increases with time, whereas the loads applied to the

soil cell will decay since the PVC pipe undergoes stress relaxation (Bishop, 1981).

In 1981, Gaube and Miller designed a sand box with 5mm thick sheet steel to test

plastic sewer pipe. They placed a 'water-filled bladder' between the tops of the soil and









the soil box lid to apply pressure to the top of the soil. Their design employed a box

width of eight times the diameter of the pipe. The soil box tests produced results similar

to field tests and indicated that plastic pipe could withstand up to 150 feet of earth fill,

assuming that compacted soil was used for the fill (Gaube, 1981).

Molin tested flexible 200mm diameter PVC pipes in a soil box, while investigating

backfills of sand, compacted clay and uncompacted clay (Molin, 1981). His results

showed that the vertical soil pressure above the pipe increases with increasing pipe

stiffness. Molin stated that his soil box test results were similar to field tests yet claimed

the field tests were more vague. Measured strains were compared to calculated strains

and found to be acceptable.

In 1985, Singhal and Veliz believed that the boundary conditions and edge effects

could be eliminated (Singhal, 1985). In other words, the soil surrounding the pipe in the

soil box carried stresses and strains that dissipated laterally with distance. Singhal and

Veliz tested cyclic torsion, axial pullout, and bending on buried pipes. In the same year

Todres and McClinton used a soil box constructed of 19mm plywood panels pinned

through steel channels and angles. A 4-inch diameter steel pipe was tested for

performance by measuring strains on the pipe's walls. A controlled load placed on top of

the fill allowed them to compare the measured bending stresses with calculated stresses,

and obtained a reasonable correlation (Todres, 1985).

A finite element analysis was implemented into the design of a soil box test facility

(Zanzinger, 1995). Zanzinger and Gartung based their design on the drop in modulus of

elasticity of the pipe over time. When loading the surface of the soil, stresses from the

pipe are distributed to the soil surrounding the pipe. A finite element analysis approach









was used to determine the required soil box width needed to eliminate stresses acting on

the sides of the box. During the 1000-hour test, a laser was used to measure pipe

deformation.

The size of a testing facility determines whether one needs to design a laboratory

test site to simulate field conditions. Some research centers are equipped with full-scale

facilities that reproduce results that would be expected in the field because the research is

essentially performed in a field test site. The Center for Pipes and Underground

Structures was developed by both the Ohio University and ORITE (Ohio Research

Institute for Transportation and the Environment) and is shown in Figures 3 and 4. This

facility is one of the largest existing test facilities of this nature. When a large test site

like this is not available, a smaller facility or soil box is needed to recreate tests under

laboratory conditions. Laboratory tests provide better control of the test and conditions.

As stated earlier, Ohio State University is home to one of the largest in situ pipe

testing facilities. Smaller test facilities are needed for researchers unable to gain access

to such large test pits. Brachman et al. states that boundary conditions, such as the

geometry of testing facilities and the method of load application, may significantly affect

test results. Advantages of soil box testing include better control and access for

instrumentation.



























Figure 2.3: Ohio University Full Scale Testing Site (Courtesy of James Hardie Pipe,
Inc.)


Figure 2.4: The Center for Pipes and Underground Structures Test Facility at Ohio
University. (Courtesy of James Hardie Pipe, Inc.)









Brachman et al. conducted research entitled "Interpretation of Buried Pipe Test:

Small-Diameter Pipe in Ohio University Facility." A small diameter leachate collection

pipe was analyzed using two and three-dimensional analyses. Numerical analysis

provides one way to assess boundary conditions on measured results when laboratory

tests are conducted. Tests were performed at the Ohio University facility for small

diameter high-density polyethylene leachate collection pipes. Boundary conditions of the

test facility, along with the stress states in the soil and the pipe's response to the soil

interaction, were investigated.

Two large hydraulic cylinders were used to apply a vertical force to a loading

platform, thus loading the soil and underlying pipe. The backfill surrounding the pipe

consisted of crushed stone over clay bedding to simulate a leachate collection system.

The most important boundary condition for this study was the method of load application.

A platform of eight W-shaped steel beams welded together provided a rigid footing for

the applied load. Results from the finite element analysis clarified the state of stress in

the soil due to the overburden load induced by the platform loading. The facility results

were compared to the expected results from the field installation.

The tests results were complex and required careful interpretation before drawing

any conclusions concerning pipe performance for leachate collection in landfills. The

stresses induced by the rigid platform differed from the expected uniform load in a

landfill. It was found that, at low load levels, a portion of the backfill material at the soil-

pipe interface yielded due to the crushed stone behaving as a beam in bending. This

effect caused a reduction in the lateral support provided to the pipe, thus increasing pipe

deformations and altering the mode of pipe deflection. The measured deflections were









higher than expected in a landfill situation. Care must be taken when interpreting the

results from facility testing. Numerical analysis can be successful when correctly

interpreted to evaluate boundary conditions.

James Hardie tested a rigid soil box constructed with lateral sidewalls of 9mm

angle iron. The angle iron effectively restricted any lateral movement. Hardie's rigid

pipe testing apparatus is shown in Figures 2.5 and 2.6.

After realizing that the rigid box was restrictive and did not accurately represent in

situ conditions, Hardie developed a flexible soil box (shown in Figure 2.7). The new box

was designed with moving lateral walls. Leaf springs supported the walls in an attempt

to simulate in situ stiffness. The flexible soil box provided the option of changing the

sidewall's lateral stiffness to reflect different burial conditions.






















Figure 2.5: Hardie Pipe's Rigid Soil Box Front View (Courtesy of Hardie Pipe Inc.)






























Figure 2.6: Hardie Pipe's Rigid Soil Box Full Image View (Courtesy of Hardie Pipe)


Figure 2.7: Hardie Pipe Flexible Soil (Courtesy of Hardie Pipe, Inc.)









Testing using this type of flexible soil (performed by Hardie Pipe) was observed on

November 13, 2002 at the University of Central Florida (UCF) in Orlando, Florida as

seen in Figure 2.8. This research will compare the soil box test results to in situ results.

Two different types of reinforced concrete pipe were tested; Hardie Pipe's fiber

reinforced concrete pipe (in both dry and saturated conditions) and standard reinforced

concrete pipe. The backfill used in these tests was compacted coarse sand. Two sheets

of Teflon were used to line the inside walls of the box in order to reduce the sidewall

friction angle to less than 5 degrees. The box was loaded using a concrete slab placed on

top of a series of parallel 2 x 4's in order to simulate a uniformly distributed load. The

loading was controlled and monitored using a personal computer.

^^^^ ^^--f 7C \m. ^ 1---^^^


Figure 2.8: Hardie Pipe Flexible Soil Box Testing conducted at UCF.









During the test, LVDT's were used to measure the deflection of the sidewalls

induced by the applied loading (Figure 2.9). The load level was noted at the point where

the first crack visible to the naked eye formed and loading was then continued until

multiple cracks appeared. The flexible soil box is equipped with a viewing port of plexi-

glass to monitor the pipe during loading. Interior surfaces of the crown and invert of the

pipe are visible from the viewing port. Cracking of the fiber reinforced concrete pipe is


Figure 2.9: LVDT's Shown to Measure Deflection of Sidewalls.

shown in Figure 2.10, as viewed through the viewing port of the soil box. Each of the

tests was continued until cracks propagated across the crown and invert of the pipe. For

both the dry and saturated conditions of fiber reinforced concrete pipes, the pipe was

loaded until cracking occurred in flexure.

































Figure 2.10: Flexure Crack At The Crown of Fiber Reinforced Concrete Pipe.

Standard reinforced concrete pipes were also tested using the flexible soil box.

Compaction of the backfill was performed in accordance with the same procedure as used

in the fiber reinforced concrete pipes. Cracking of the reinforced concrete pipe occurred

at the crown of the pipe, as shown in Figure 2.11. Though the same loading procedure

was used for the reinforced concrete pipes, it resulted in a much lower maximum load

when compared to the fiber reinforced concrete pipes. The testing observed on

November 13, 2002, was only a portion of the testing program being conducted at the

University of Central Florida.

































Figure 2.11: Reinforced Concrete Pipe Cracking At The Crown.

Brachman et al. discussed the design of a laboratory facility and the testing of

buried pipe performance. Their study considered a limiting applied pressure of 1000kPa,

based on a burial length of 50 m and a soil density of 20kN/m3, and then used finite

element analysis to determine the effect of sidewall friction on the soil. Symmetry about

the vertical diameter of the pipe was assumed so that only one half of the test facility

need be modeled. The load was applied as a uniform pressure and the soil box contained

a 2,000 mm2 soil prism that extended to a height of 1,600 mm. These dimensions

allowed only small horizontal deflections under large vertical pressures. The large

distance between the pipe and the sidewalls was an attempt to provide lateral stiffness,

while at the same time avoiding alteration of the pipe's behavior.

Hardie stated that the effect of sidewall friction should be considered with respect

to the pipe's response and not to the soil box wall. Brachman et al. stated that sidewall









friction reduces the amount of load experienced by the pipe, thus resulting in a reduction

in vertical deflection of the pipe. Based on a finite element analysis, Brachman et al.

concluded that a sidewall friction angle of 5 degrees best simulates in situ conditions. In

order to obtain a friction angle of 5 degrees, Brachman used polyethylene sheets

lubricated with DC44 silicone grease. His research highlighted the importance of

recognizing that the pipe distributes both horizontal and vertical stresses and that a

reasonable model of the soil stresses can be achieved when the top and bottom of the soil

is at least a distance of one diameter from the pipe. Research conducted at the University

of Florida will provide a bedding depth below the pipe of one pipe diameter.

In the early years of soil box testing, the boundary conditions induced by the

equipment caused poor correlation with in situ condition results. Boundary conditions

were later revised to better represent such conditions. From the literature review

conducted, it has become apparent that boundary conditions and pipe installation

technique are of high importance when trying to simulate in situ performance with pipes

tested in a soil box. Therefore when designing a soil box, controlling boundary

conditions and soil stiffness is fundamental to producing accurate and acceptable results.














CHAPTER 3
STANDARD REINFORCED CONCRETE PIPE VS. FIBER REINFORCED
CONCRETE PIPE

Background Information on Concrete Pipes

Presently, concrete pipes are fabricated in a variety of sizes, ranging from 4 inches

to over 16 feet of inner diameter. Concrete pipes are used to transport liquids under

gravity flow and are implemented as highway culverts, storm drains and sanitary sewers.

The evaluation of concrete pipes for use as storm drains under gravity flow is the focus of

this project.

Concrete pipes are reinforced against crushing when the inner diameter is greater

than 24 inches. Standard reinforced concrete pipes (SRCP) are fabricated in accordance

with ASTM C76 "Standard Specification for Reinforced Concrete Culvert, Storm Drain,

and Sewer Pipe." ASTM C76 covers a size range diameter from 12 inches to 108 inches,

with an exception for larger pipe diameters. SRCP are heavy products and require lifters

capable of proper placement and installation. The SRCP used for this research will come

from Rinker Materials.

Hardie Pipe introduced fiber reinforced concrete pipes (FRCP) into the civil

construction market for large drainage pipes in early 2002 under the company's

trademark Fiber Reinforced Concrete Speed Drain Pipes.

The concrete properties for each type of pipe include compressive strength, density,

absorption, water-cement ratio, cementious materials, aggregates and versatility.









Cracking in Concrete--The Fracture Zone

Cracking in concrete begins at the micro level. The fracture zone is defined as the

state when the stress or the strain is increased to the point where the atomic bonds within

the matrix are broken and the solid material is cracked or fractured. As the stress load

increases, cracks will develop and propagate. In concrete, cracks typically propagate due

to tensile stresses, as a result of the low tensile strength of concrete. Though tensile

strength is not considered in concrete design, is the cause of most crack initiation and

propagation. Failure of concrete is due to tensile stresses induced by loads and/ or

environmental changes. Concrete failure is often the end result of microcracking

associated with the interfacial region between the hydrated cement paste and aggregates

or other inclusions (reinforcing steel, fibers, etc.). Cracks are initially localized but

increase in size as the applied stress is increased. In certain circumstances, cracks can

propagate very fast. Cracks propagate in three different modes.

Cracking modes are classified as either plane strain modes or anti-plane strain

modes. Mode I and Mode II deformation are plane strain and Mode III is anti-plane

strain. The deformation of the crack is discussed using the coordinate system shown for

Mode I displacement in Figure 3.1 below. A description of the three modes of cracking

will follow the coordinate system shown in Figure 3.2. In Figure 3.1, an isotropic solid is

shown with the origin of the coordinate system at the tip of the crack. It is important to

note that the solid shown in Figure 3.1 represents an isotropic solid only. Anisotropic

solids are quite complicated when analyzing the fracture of the solid.

Mode I deformation occurs when a transverse plane stress is applied to the crack

forcing the crack to open up along the y-axis. Mode I is the most prevalent type of

cracking in a brittle material. Mode II deformation is a result of a shear stress applied on











the cracked solid that forces the faces of the crack slide over one another parallel to the

xy plane. Mode III occurs when a shear stress is applied resulting in the crack faces

sliding over each other perpendicular to the xy plane. Cracking of concrete occurs in

different stages beginning at the micro level.






T,-

r(





/ ^ Leading edge
S\of the crack









Figure 3.1 Coordinate system and stress components ahead of crack tip (Mode I
displacement) (Mindess, 2003).


Mode I Mode II
opening mode sliding mode


Figure 3.2 Three Modes of Cracking (Mindess, 2003).


Mode III
tearing mode









Standard Reinforced Concrete Pipe--SRCP

SRCP is widely used in the construction market today. Uses include highway

culverts, storm drains and sanitary sewers. Pipes without reinforcement are used where

the application is suitable for such products. The majority of applications require

reinforcement in order to increase overall strength and resist loads applied over the

service life of the pipe. For example, a highway culvert located under a bridge would be

subjected to cycling of live loads from traffic. Reinforcement would increase the amount

of live load capability prior to failure.

SRCP uses steel rebar for reinforcement. The rebar is oriented within the pipe as a

longitudinal spiral or linear section placed around the perimeter, parallel to the length of

the pipe. Welded wire mesh is also used for additional reinforcement, allowing for

increased bonding of the concrete to the rebar reinforcement. An SRCP is shown in

Figure 3.3 with the steel reinforcement visible. This pipe section was tested in Hardie

Pipe's soil box apparatus located at the UCF. The aggregates are also visible, along with

the cracking induced by the load applied.

Properties that will be discussed for SRCP include durability, such as the ability to

resist corrosion of rebar and proper bonding of the concrete to the rebar. Without

sufficient bonding of the concrete and rebar, the reinforcement cannot properly carry the

tensile stresses and prevent cracking of the concrete matrix.

Mechanics of SRCP

Different sizes of steel rebar are used to reinforce cement based concrete structures.

In this research, concrete pipes will be tested for ultimate strength. Concrete is weak in

tension, strong in compression. When tensile stresses are induced in the










































Figure 3.3 Standard Reinforced Concrete Pipe Section Cracked.

concrete, the use of steel bars or wires carries the tensile stresses. Sufficient bond

between the concrete and steel is necessary to allow transfer of the tensile stresses to the

steel. The structural performance of SRCP thus relies on the bond between the steel and

concrete. Steel design for concrete structure is specified in ACI 318.

Steel rebar left unprotected from the environment will corrode, resulting in a loss of

strength. In reinforced concrete structures, the steel rebar is placed within the concrete

matrix to provide protection from corrosion.









Reinforcement

The most commonly used reinforcements for non-prestressed members are hot-

rolled deformed bars and wire fabric. Hot-rolled deformed steel bars are basically round

in cross section with lugs or deformations rolled into the surface to aid in developing

anchorage or bond with the concrete. Steel reinforcing bars are manufactured according

to ASTM specifications: ASTM A615-85 Specification for Deformed and Plain Billet-

Steel Bars for Concrete Reinforcement; ASTM A616 Specification for Rail-Steel

Deformed and Plain Bars for Concrete Reinforcement; ASTM A617 Specification for

Axle-Steel Deformed and Plain Bars for Concrete Reinforcement; ASTM A706

Specification for Low-Alloy Steel Deformed Bars for Concrete Reinforcement.

Strength

Reinforcing steel bars are manufactured in three grades; 40, 50, and 60, with yield

strengths of 40, 50 and 60 thousand pounds per square inch (ksi). The 40-ksi bar is the

most ductile of the three. Hot rolled steel bars are shown in Figure 3.4. Steel

reinforcement is classified by its nominal diameter, expressed in eights of an inch.

ASTM specifications for reinforcing bars define the yield strength as the stress

carried by the bar at a strain of 0.005. ACI Sections 3.5.3.2 and 3.5.3.4 to 3.5.3.6 define

the yield strength as the stress carried at a strain of 0.0035. ACI's definition is based on

the strain at which concrete crushes when bars are in compression, a situation where a

strain of 0.005 may never be reached. ASTM specifications for yield strength are based

on mill tests that are carried out at a high rate of loading.




















Figure 3.4 Standard Reinforcing Steel Rebar.

Structural Performance

Structural performance is affected by several factors, such as bond, temperature and

surrounding environmental conditions. The bond between the concrete matrix and the

steel defines the stress transfer within the concrete. Tensile stresses are absorbed by the

structural steel, providing tensile strength to the concrete structure. Debonding of the

concrete and steel results in lower overall strength and performance of the concrete.

Temperature affects the strength and performance of steel. The concrete cover over the

steel resists heat penetration into the concrete, preventing temperature rise and decreased

strength in the steel. Steel temperatures exceeding about 8500F result in a significant

drop in both yield and ultimate strength.

Adverse environmental conditions can cause corrosion of the steel reinforcing bars,

thus decreasing performance. Water and oxygen penetration into the concrete from the

surrounding soil or water increases the possibility of steel corrosion, especially in the

presence of chlorides. Concrete pipe serving as storm drains are susceptible to water

penetration from both inside and outside the pipe. Increasing strains applied to the pipe

will induce stresses and result in cracks forming throughout the pipe. Once a crack

occurs, channels are created throughout the concrete, allowing water to enter at a much









higher rate and accelerating corrosion of the steel reinforcement. Once the steel corrodes,

the strength of the concrete structure is impaired.

Fiber Reinforced Concrete Pipe--FRCP

The fibers in fiber reinforced concrete pipe serves as the reinforcement used to

carry tensile stresses. A diameter range of 12" to 48" is produced, with a standard length

of 16'. Pipe strengths within each size category are divided into five different classes,

based upon wall thickness (I V, with I being standard and V extra heavy). FRCP is

manufactured in accordance with ASTM Standard C1450 and FDOT Standard 941, with

designs from AASHTO Section 17 or LRFD Section 12. A typical FRC pipe is shown in

Figure 3.5.


Figure 3.5 Typical Fiber Reinforced Concrete Pipe









Mechanics of FRCP

Different types of fibers are used to reinforce cement-based matrices. Fibers offer a

more economical means of reinforcement for concrete structures of small size.

Economically, fibers are cheaper than the usual steel reinforcing bars used, although

fibers cannot be considered a replacement for traditional structural reinforcement of

massive structures. The ACI 318 design code is based solely on concrete strength as the

design criteria, considering only the peak loads a structure can withstand. Fibers are used

very little in conjunction with structural steel due to the lack of consideration for post

peak behavior. Fibers are most effective in the post peak phase of loading, since the

fibers do not carry significant load until after the first crack occurs. They thus do no

affect the first-crack strength of concrete and have very little effect on ultimate load.

Fibers are generally randomly distributed throughout the cross section. Most fibers

are short and closely spaced within the matrix to allow formation of a bond between the

fiber and the matrix. Stresses are sometimes transferred from fiber to fiber within the

matrix, thus resulting in fiber-fiber interaction.

Fibers are primarily used to control crack propagation by bridging across cracks as

they begin to open when the concrete strain exceeds its ultimate capacity. An important

factor used to determine fracture properties of a material is the stress intensity factor.

Cracking occurs after the critical value of the stress intensity factor is reached. Without

going into the detail necessary to solve for the stress intensity factor, a simple overview

will be provided to explain its importance.

The stress intensity factor K is considered to be a single-parameter description of

the stress and displacement fields in the region of a crack tip (Mindess, 2003). When the









stress intensity factor reaches a critical value, unstable fracture will occur, resulting in

cracks forming.

Fibers contribute to a more ductile concrete matrix. Fracture toughness is increased

due to the bridging of fibers across the crack as they resist crack opening. As the applied

load is increased, fiber reinforcement is designed to absorb the maximum amount of

energy possible before unstable behavior of the matrix occurs.

Fiber-Matrix Bond

In properly designed cement composites for maximum performance, fibers are

randomly dispersed throughout the matrix. The fiber matrix bond is a function of both

the fiber and matrix properties. Fiber-matrix bond strength is determined from fiber

pullout tests and reported as an average value over fiber surface area. Stresses are

induced on the fiber-matrix bond, resulting in fiber pullout or debonding from the matrix

after the maximum strength is reached. Typical fiber-matrix pullout strengths are shown

in Table 3.1.

When using cement as a matrix, the fiber-cement interface can become complicated

if a chemical reaction occurs between the cement and fiber. Formation of water around

the fibers can occur due to bleeding in fresh concrete or insufficient packing of cement

grains around the fibers. Under such conditions, the matrix becomes porous near the

surface of the fibers than in the bulk cement paste.

Fiber-Fiber Interaction

A high ratio between fiber modulus of elasticity and matrix modulus of elasticity

facilitates stress transfer from the matrix to the fiber. Fiber to fiber interaction occurs

when stress is transferred between fibers. Fibers take on the tensile stress induced within

the fibrous composite material. Stress concentrations will arise at the fiber ends if they









are discontinuous. The tensile stress that would be assumed by the fiber without the

discontinuity must be taken up by the surrounding fibers in the composite.

Table 3.1 Typical Fiber-Matrix Pullout Strengths (Mindess, 2003).

Matrix Fiber Pullout Strength Mpa (lbs/in2)
Cement Paste Asbestos 0.8-3.2 (115-460)
Glass 6.4-10.0 (930-1450)
Polycrystalline alumina 5.6-13.6 (810-1970)
Steel 6.8-8.3 (990-1200)
Mortar Steel 5.4 (780)
Concrete Steel 3.6 (520)(first crack)
4.2 (610)(failure)
Nylon 0.14 (20)
Polypropelene 1 (150)


The effect of fiber-fiber interaction on stress transfer is described by Riley's

theory (Beaudoin, 1990), which states that discontinuous fibers can contribute a

maximum of only 6/7 of their strength to the strength of the composite, decreasing to 1/2

for badly flawed fibers.

Load vs. Deflection in Fiber Reinforced Concrete

As the applied load on a material increases, the strain will increase as well resulting

in deformation and deflection. A typical stress-strain curve for fiber reinforced concrete

is shown in Figure 3.6. Point A indicates load at which the first crack occurs in the

matrix, known as the first-crack strength. The stress at which the first crack occurs is the

same in fiber reinforced concrete, traditionally reinforced concrete and plain concrete.

The strength of fiber reinforced concrete in the post-cracking zone comes from the

transfer of loads across the cracks by the fibers, increasing the strength of fiber reinforced

concrete over that of the matrix. Fibers increase the toughness by providing energy

absorption mechanisms through the gradual debonding and pull out of the fibers bridging

across the cracks (Mindess, 2003).
























Deflection in flexure
Figure 3.6 Typical load-deflection curve for fiber reinforced concrete in flexure
(Mindess, 2003)

Stages of Cracking-Fiber Intervention

The first stage of cracking is the development of microcracks. Microcracks form

due to high stress concentrations in specific regions of the concrete matrix. The

microcracks lengthen, meet and coalesce to form one or more macrocracks. The final

stage corresponds to the propagation of cracks. Much research has been done to find out

how fibers intervene during the various stages of cracking. During the first stage, the

fibers respond to the uniformly distributed microcracking creating a stitching effect on

the micro-cracks and preventing propagation. The intervention of the fibers retards the

microcracking coalescence phase and the creation of macro-cracks. Once macrocracks

develop, the fibers will bridge across them and serve as reinforcement similar to the steel

in reinforced concrete. This bridging mechanism is illustrated in Figure 3.7.









Load. P

Traction-free crack length








interlock



Load. I'
Fiber bridging length
Figure 3.7 Schematic representations of fibers bridging across a crack (Mindess, 2003).

Fibers bridge across the cracks as they open within the matrix. The stress field

around the crack is shown along with the traction-free crack length, fiber bridging length

and the aggregate interlock. Stresses are absorbed by the fibers in three different areas

shown in figure 3.7. The traction free zone is where fibers have pulled out due to the

crack opening up wide enough to overcome the pull out strength. Stresses are absorbed

by the fibers and transferred across the fibers by frictional slip in the fiber-bridging zone.

Aggregates also absorb energy as their interlocking is distributed to the matrix itself in

the microcracked matrix process zone.

As stated previously, fibers intervene at two levels; the material level (during

macrocracking) and at the structural level (during stress redistribution). In order for the

fibers to respond effectively, fiber dimensions and properties must be optimized for the

matrix material. The volume proportion of fibers also needs to be optimized to provide

maximum mechanical performance. Two approaches are considered during the mix

design stage. Since only discontinuous fibers are used, one possibility is to mix in a high

percentage of relatively short fibers, resulting in an increase in the strength of the FRC, as

the dimensional scale of the fibers is the same as that of the microcracks. Another









possibility is to mix in a low percentage of long fibers, resulting in improved ductility of

the structure. This is effective, with regard to macrocracking, since the fibers have

sufficient anchorage length on either side of the crack as it opens across the fibers.

Strength

The role of fibers in concrete is not to increase the overall strength. Some minimal

increase in strength will occur in compression and flexure. Research shows that in direct

tension, where it would be expected that fibers should be most effective in terms of

strength, strength increases are limited to about 30% for a steel fiber volume of 1.5%

(Mindess, 2003). Fibers are reported to have no major effect on shear and torsional

strength or on elastic modulus.

Toughness

Fibers have an enormous effect on toughness. If the fibers possess sufficient

strength and stiffness, and bonded well with the matrix, they will minimize cracking.

This allows the fiber reinforced concrete to withstand significant stresses over a relatively

large strain capacity in the post-cracking (or strain-softening) stage, thus providing a

considerable amount of post-cracking ductility (Mindess, 2003). Certain fibers have a

greater effect on increasing the toughness of fiber reinforced concrete. Deformed fibers

tend to have a greater effect on increasing toughness since they bond to the matrix better,

increasing the overall pullout strength. Increasing the bond strength past the point where

the fibers themselves begin to fail in tension, however, is counterproductive. More

energy is required to pull a fiber out of the surrounding matrix than to actually break the

fiber across a crack. Thus, designing fibers and FRC for fiber pullout, instead of

breakage, is the key to maximizing energy absorption and ductility during failure.









Mindess shows that steel fibers are more effective that polypropylene fibers in improving

toughness because of their higher stiffness.

FRCP versus SRCP

Performance of a pipe is based on its ability to maintain shape under service loads,

resist cracking from applied stresses, and resist deterioration of the pipe material due to

environmental exposure. Concrete drainpipes are subjected to a number of deterioration

mechanisms. Stresses will induce microcracking that eventually coalesces into

macrocracks within the concrete matrix. These cracks provide openings for water and

other deleterious materials to penetrate into the concrete. In traditional steel reinforced

concrete pipe, such ingress can induce corrosion of the steel reinforcement, leading to a

decrease in the ultimate strength of the pipe. Though not susceptible to corrosion, the

durability concerns related to cellulose fiber reinforced concrete pipe have not yet been

fully investigated. A comparison of the properties and characteristics of FRCP and SRCP

will be discussed.

Manufacturing

FRCP is manufactured using high-pressure autoclaving, allowing higher concrete

strengths to be obtained for a given set of ingredients. SRCP is manufactured with a low

water to cement ratio concrete mix, which is cast into forms during an automated

fabrication sequence.

Installation

The unit weight of a concrete pipe has a significant impact on the installation

process. A lighter pipe is easier and quicker to install than a heavy pipe. For a given size

and class, an FRC pipe is about half the weight of a corresponding SRC pipe, making it

easier to install and handle.









The standard length of SRCP is six feet. FRCP, on the other hand, comes in a

standard length of 16 feet. The installation work required for FRCP, compared to SRCP,

is cut in half due to this longer standard length. FRCP is also easier to cut when length

adjustments are needed, requiring less time than SRCP.

An important part of the installation of concrete pipes is the joint seal at the ends of

the pipes. Both pipes use rubber gaskets at joints to provide a tight seal, preventing the

flow of liquid into, and out of, the pipe system.

Performance

A vital property affecting mechanical performance is the bond of the concrete to

the reinforcement. Steel rebar is more difficult to bond to than cellulose fibers, which are

randomly dispersed throughout the concrete mix. This random dispersion allows for

increased bonding due to the small size of the fibers and the increase in total surface area

inherent in smaller particle sizes. Even with these improvements in bond performance, it

still must be remembered that fiber reinforcement cannot be considered a complete

replacement for the reinforcement of concrete structures.

Both FRCP and SRCP will be tested in the soil box designed by the University of

Florida for research funded by the Florida Department of Transportation. The FRCP will

come from Hardie Pipe, Inc. while Rinker Materials, Inc will provide the SRCP.

Following design and construction of the soil box testing apparatus, each type of pipe

will be evaluated as to its insitu performance.














CHAPTER 4
INSTRUMENTATION AND TESTING TECHNIQUES

Crack Detection and Deflection in Concrete Pipes

During the testing, the two primary parameters, other than applied load, that will be

monitored include cracking and deflection. The ultimate failure of a pipe specimen will

eventually be define relative to one or both of these serviceability limits. There are many

test methods used today to detect cracks and measure deflection of a structure under load.

Determining which of these to implement comes down to applicability, accuracy, and

ease of use.

Since the pipes are to be tested in a buried condition, detecting cracks occurring on

the outer face of the pipe may prove problematic. With current advances in technology,

crack detection is more easily accomplished using acoustic emission monitoring

techniques. This should allow detection of crack formation without the need to actually

observe the crack.

The deflection of the concrete pipes is an observation of the pipe's deformation

while subjected to an applied load. A novel method will also be used to observe and

record pipe deflection during testing, specifically three-dimensional imaging of the inner

surface of the pipe with a laser based digitizing system.

Background on Acoustic Emission Testing

Acoustic emission is defined as an acoustic wave generated by a material when

subjected to an external stimulus causing an irreversible change in the material. There

are two types of acoustic emission signals, continuous and burst signals. A continuous










emission is a sustained signal level produced by rapidly occurring emission events such

as plastic deformation. A burst emission is a discrete signal related to an individual

emission event occurring in the material, such as a crack forming or propagating in a

brittle material, such as concrete.

An acoustic emission burst signal is shown in Figure 4.1. The term "acoustic

emission signal" is often used interchangeably with simply "acoustic emission". An

acoustic emission signal is defined as the electrical signal received by the sensor in

response to the acoustic wave moving through the material. This emission is picked up

by the sensor and transformed into an electrical signal, then analyzed by acoustic

emission instrumentation, resulting in information about the material that generated the

emission.

1Re Tm Duration (D)
Rise Time (R)


S Amplitude (A)


TV Tirm



Counts


MARSE, Energy Counts (E)

I Tinie
--- ----- --------- ^ T m ie


Figure 4.1: Burst Acoustic emission signal with properties

Acoustic emission is a passive, non-destructive monitoring technique. This means

that there is no input from an outside source; the technique purely monitors the material

being tested. Acoustic emission is typically used to detect cracking, delamination (slip

between concrete and steel reinforcement), failure of strands in prestressing tendons, and










fracture or debonding of fibers in fiber reinforced concrete. A typical acoustic emission

system setup is shown in Figure 4.2.



Acoustic Emission Process

Signal
Sensor



Stimulus Measurement
Recording
= 110 Interpolation
Source Acoustic I Evaluation
e Emission
Wave

Acoustic Emissions are transient elastic waves
generated by the rapid release of energy from
localized sources within a material


Figure 4.2: Acoustic Emission Process

Equipment and Instrumentation

An Acoustic emission system includes at least one sensor and a preamplifier. Most

systems also include postamplifiers and signal processors. The system to be used in this

research project is a LAM a Local Area Monitor that consists of eight sensors, which

will be mounted on the pipe during load testing. More specialized equipment often

associated with such a system includes transient recorders, spectrum analyzers,

distribution analyzers and spatial discrimination circuits (Beattie, 1993). Microprocessor

based systems have become more widely used in recent years that can perform single

channel analysis along with source location for up to eight AE channels.

Sensors

The Acoustic emission sensor is the most important part of the instrumentation and

must be properly mounted to assure the required sensitivity. Selection of sensors and or

transducers will depend on test parameters and desired results. Sensors are calibrated









using test methods stated by societies. ASTM states that annual calibration and

verification of pressure transducer, AE sensors, preamplifiers (if applicable), signal

processor (particularly the signal processor time reference), and AE electronic simulator

waveformm generator) should be performed. Equipment should conform to

manufacturer's specifications. Instruments should be calibrated in accordance with

National Institute for Standards and Technology (NIST) specifications.

An AE electronic simulator, used in making evaluations, must have each channel

respond with a peak amplitude reading within + 2dBV of the electronic waveform output.

A system performance check should be done immediately before and after an AE

examination. The preferred technique for this is the pencil lead break test. A complete

description of this test can be found in ASTM E 570.

One very important factor affecting sensor performance is location. Determination

of the number of sensors required for the test, their placement strategy and location on the

specimen to be monitored is critical. A single sensor used near the expected source of

AE is sufficient when background noise can be controlled or does not exist. When

background noise is limited, the use of a single AE data sensor near the expected source

plus a guard sensors) near the background source will suffice. ASTM defines a guard

sensor as sensors whose primary function is the elimination of extraneous noise based on

arrival sequences.

Another technique involves the placement of two or more sensors to perform

spatial discrimination of background noise and allow AE events to occur. ASTM defines

spatial discrimination as the process of using one or more (guard and data) sensors to

eliminate extraneous noise based on arrival sequences.









In situations where irrelevant noise cannot be controlled during testing and could

be emanating from any and all directions, a multiple sensor location strategy should be

considered. Using a linear or planar sensor configuration will allow for accurate source

location of the acoustic emission event. Applications of spatial filtering and/or spatial

discrimination will only allow data emanating from the region of interest to be processed

as relevant AE data.

Preamplifiers

Preamplifiers (as shown in Figure 4.3) are used to prevent loss of sensor activity.

Loss occurs when one sensor is connected through a long coaxial cable to an amplifier.

The amplifier is split into a fixed gain preamplifier located close to the sensor. The

preamplifier consists of a low noise input stage, bandpass filters and a low impedance

output stage capable of driving a 50-ohm cable (Beattie, 1993). Power for the

preamplifier is received from the main instrument group. AE preamplifiers are designed

to have a relatively flat frequency response between about 20 kHz and 2MHz, without the

bandpass filters (Beattie, 1993).

Preamplifiers can be included in the sensor package. The advantages of this

arrangement are the elimination of cable capacity effect and being able to tailor the

preamplifier characteristics to match the sensor. The disadvantages of such units, besides

their higher cost, are that they are restricted to temperatures near 200C (the preamplifier

will not work properly at higher or lower temperatures) and that a separate preamplifier

has to be purchased for each sensor (Beattie, 1993).

Postamplifiers and Signal Processors

Most AE systems use variable gain postamplifiers. This allows the use of signal

processors with fixed input ranges or thresholds in conjunction with fixed gain









preamplifiers (Beattie, 1993). The systems total gain is the sum of the preamplifier and

postamplifier gains, expressed in decibels. Additional noise reduction can be achieved

through postamplifiers from bandpass filters.

Signal processors are typically included in the system's capabilities. These include

voltage controlled gates that allow data to be collected only on certain portions of a load

cycle, envelope processors which attempt to filter out high frequencies leaving only the

signal envelope to be counted, logarithmic converters which allow the output of the

signal analyzer electronics to be plotted in logarithmic form and a unit which allows the

combination of outputs from several preamplifiers so that several sensors can be

monitored by one channel of electronics (Beattie, 1993).













Figure 4.3: Pre-amplifiers (PAC)

Transient Recorders

Transient recorders are used to study individual AE burst signals. A signal is

digitized in real time, and then stored into memory. A transient recorder is used in

sequence with an oscilloscope or spectrum analyzer to display AE signals at visible

speeds. Digital rates vary on transient recorders. The fastest rate of the recorder is

ultimately the limiting rate, with some instruments sampling up to 1 word/nano-second

(Beattie, 1993).










Sampling rates can be modified for testing purposes. One advantage of transient

recorders is an additional mode of triggering and pretriggering, where the input signal is

continuously digitized and the data stored in the memory (Beattie, 1993). This feature

allows a digitized picture of the signal to be displayed as it is received. More advanced

systems allow recording of two or more signals simultaneously. The recording of more

than one AE signal is shown in Figure 4.4.


#58 channel 1









-2 58 2ch0 nneOO
N68 channel 4

N58 channel 5 A

N58 channel -



N#58channel 8

-250 1 250 511 750
timre[j s]
Figure 4.4: Transient recorder with multiple AE signals

Spectrum Analyzers

The ideal spectrum analyzer has horizontal and vertical signal outputs, allowing an

X-Y plot of the AE signal. Spectrum analysis is done either directly with analog

electronics or by digitally viewing a continuous signal. A local oscillator signal is mixed

with the input signal and the highest frequency is passed through a chain of intermediate

frequency (I.F.) amplifiers, after which it is measured by a voltmeter (Beattie, 1993).

The oscillator is swept through a frequency range so that the frequency components of

the signal selected by the I.F. amplifier are continuously changing (Beattie, 1993). The









voltmeter output is plotted on the vertical axis of the oscilloscope. A typical oscilloscope

is shown in Figure 4.5. A synchronized signal is displayed for the horizontal axis

through the local oscillator frequency. The final result is a plot of signal strength vs.

frequency.

Spectrum analyzers range in frequencies from at least 10kHz to 2MHz. The width

and speed of the local oscillator and the sharpness of the I.F. amplifier filters are all under

the direct control of the operator (Beattie, 1993). An AE burst signal emission is not

suitable for spectrum analysis. AE burst signals must be captured on a tape recorder and

played repetitively onto the spectrum analyzer for analysis. This essentially simulates a

continuous signal for spectrum analysis.

















Figure 4.5: Digital Oscilloscope

LAM--Local Area Monitor

LAM is the world's first acoustic emissions system to allow remote condition

monitoring of structures. Physical Acoustics Corporation developed the LAM, as shown

in Figure 4.6, in conjunction with the U.S. Federal Highway Administration. The system









is portable and easy to handle weighing only 25 pounds including one battery pack.

Features of the LAM used in this research project are:

Modular 8 channel DSP-based AE system with 16-bit A/D
User-friendly software
AC/DC powered
4 high-speed and 8 low-speed parametrics
Digital AE features and waveforms processed simultaneously
Software programmable filters
Resistant to harsh environmental conditions



8-Channel
DSP LAM
4 Highly Sped
Paramatrie










a Ekinai, .Battery
EventlGII 21 Inpui '
Figure 4.6: LAM-Local Area Monitor

The LAM is useful for monitoring many structures including bridges for defects,

chemical/ petrochemical tanks for leaks and deterioration, transformers for partial

discharge and most structures (including concrete pipes) during fatigue tests.

The LAM was originally designed for monitoring defects in steel bridges. The

LAM's modularity lends itself to many other applications, as stated before, to monitor

fatigue cracks and other discontinuities in structures, pressure vessels and transformers.

This research will use the LAM to monitor microcracking in the walls of fiber reinforced

and steel reinforced concrete pipes.









The LAM offers up to 8 channels of digital AE for short-term condition

monitoring, long-term integrity monitoring, laboratory fatigue testing or incipient failure

detection monitoring through user selections. The unit operates from an external 12 Volt

DC battery or 110 Volt AC power supply. An optional feature is remote access through

traditional phone line or cellular phone. This feature allows the user to monitor the

apparatus from an office or other location remote from the test site.

Advantages of the LAM over other acoustic emission instrumentation include a

reduction in the extensive cabling normally required for operation. The unit is placed on

site with the structure/ object to be monitored. For this project, the concrete pipes will be

instrumented with the eight sensors to monitor microcracks occurring inside the walls.

Data analysis for this research will be performed with the NOESIS 3.1 software

published by Physical Acoustics Corporation.

NOESIS 3.1 is an MS Windows based advanced data analysis pattern recognition

and neural networks software used for acoustic emission applications. It provides all

necessary tools for analyzing, filtering and classifying acoustic emission hits and

waveforms that are acquired with the LAM unit. NOESIS is equipped to handle data

saved in a DTA file format produced by the LAM unit. The software utilizes PAC

(Physical Acoustics Corporation) file libraries to load and save data in the DTA file

format.

NOESIS allows multiple DTA files to be loaded simultaneously for direct

comparison, statistical analysis and filtering or merging. Direct export of data files to MS

EXCEL and MS WORD is also possible. Any number of windows can be displayed,

limited only by the resolution of the viewing window for adequate visibility. Navigation









throughout the program, along with data selection, is done with the mouse. Data point

selection is available using the mouse on scatter plots, cumulative plots, waveform plots,

FFT plots and tabular data views. NOESIS allows graphs and plots to be customized for

presentation of data and analysis.

Waveforms and AE hits can be selected and displayed on any graph or data table.

Comparison of the data can be superimposed or viewed in three dimensions. Graphical

and other data filtering are applied to waveform views for presenting the collected data.

Any changes made to hardware settings are immediately reflected in the waveforms.

Minolta 3D Digitizer--VIVID 900

Minolta is the world's largest manufacturer of 3-D non-contact digitizing

instruments, providing a 3-D scanner with a simple point and shoot camera that scans

300,000 points in less than 3 seconds. This project will use Minolta's newest 3-D

scanner, the VIVID 900, shown in figure 4.7. The VIVID 900 is an easy to use scanner

with simplicity, flexibility and portability. Minolta offers simplicity by a point and shoot

camera with excellent results. The VIVID 900 includes interchangeable lenses, for

variable scanning volumes, for flexibility. The camera unit is compact, measuring 8-

3/8"x16-1/4"x 10-11/16" and weighs only 24 lbs. Scans can be saved and stored on a

compact flash memory card or viewed immediately after scanning on the rear-panel's

color LCD viewfinder. Color images are equivalent to a 3 CCD digital camera

displaying full 24-bit color depth.

Hardware

The Minolta VIVID 900 offers variable volumes for digitizing between 110 x 80 x

40 mm and 1200 x 900 x 750 mm. There are three interchangeable lenses included as

standard accessories for scanning; telephoto, medium and wide angle. The VIVID 900 is








an independent instrument that does not require a host computer for operation. Scanned

images are saved to a flash memory card (Figure 4.8) or viewed immediately after

scanning on the LCD viewfinder. The Minolta VIVID 900 also offers an autofocus

function that eliminates the need to move the unit back and forth to achieve optimal focus

for the scan.





'S'






Figure 4.7 Minolta VIVID 900 Non-Contact 3-D Digitizer







FLASHVISK
MASS STORAGE



Sain)isk ^


Figure 4.8 Compact Flash Memory Card (40MB capacity)

Accessories

Minolta offers accessories to improve the outcome of the scan. When scanning a

complete 3600 view of an object, the rotary specimen stage shown in Figure 4.9 will










prove helpful. A rotating disc is set to rotate at a specified speed while the VIVID 900

scans the object, resulting in a full 3-D image of the object. To ensure level scans and

stability, Minolta offers a tripod accessory with the option of a tilt mounting base unit as

shown in figure 4.10. Other accessories available are a PC card adapter allowing direct

transfer of the scanned images to a personal computer for analysis.








Figre.9otaingStageetfoSca.............F .... 3 .-D.
II : I-




H'ii'






... .... .. ..... : I... ::... .:: .... .


Fiur-.9Rtaig-taeSe-orSanin--Fl-3Dimg
Figure 4.9 Rotating Stage Set for Scanning a Full 3-D image
-..4..


Figure 4.10 Tripod (left) and Tilting Base Mount (right) for Minolta VIVID 900









Operation

The Minolta VIVID 900's basic theory of operation is described through LASER

triangulation. A laser source from the VIVID 900 emits a horizontal light stripe through

a cylindrical lens onto the object being scanned. The plane of light is swept across the

field of view by a rotating mirror. The light is reflected from the scanned object,

captured, and observed by a single frame through the CCD camera. Once received by the

CCD, the light is converted through triangulation into distance information. Each scan

line is captured and observed by the CCD camera. The shape of the image of each

reflected scan line is derived to produce the contour of the surface. The selected area in

the view is captured in 2.5 seconds (0.3 seconds in FAST mode), and the surface shape is

converted to a lattice of over 300,000 vertices or connected points (Minolta, 2001).

The VIVID 900 produces polygonal-mesh with all connected information,

eliminating geometric ambiguities while improving detail. Minolta's VIVID 900 uses an

X, Y, and Z coordinate axis. The x coordinate is the horizontal dimension of the focal

plane, the y coordinate is the vertical axis and the z coordinate is the distance from the

sensor. The VIVID 900 creates no parallax error. Specifications for the VIVID 900 are

displayed in Table 4.1 below.









Table 4.1 Specifications of Minolta VIVID 900 3D digitizer.


Type
Measuring method
AF
Light-Receiving Lens
(Exchangeable)

Image Input Range
Measurement Input Range
Laser Output

Laser Scan Method
Input Time
Transfer Time to Host
Computer
Ambient Light Condition
Imaging Element

Number of Output Pixels


Output Format




Recording Medium
Data File Size

Viewfinder
Output Interface
Power

Dimensions (WxHxD)
Weight
Operating environment

Storage Temperature


Non-contact 3D digitizer VIVID 900
Triangulation light block method
Image surface AF (contrast method), active AF
TELE: Focal distance f=25mm
MIDDLE: Focal distance f=14mm
WIDE: Focal distance f=8mm
0.6 to 2.5m (2m for WIDE)
0.6 to 1.2m
"Eye-safe", Class I (FDA), Class 2 (IEC),
Maximum 30mW 690 nm
Galvano mirror
0.3 sec (FAST mode), 2.5 sec (FINE mode), 0.5 sec (FINE mode)
Approx. 1 sec (FAST mode, 1.5 sec (FINE mode)

Office Environment, 500 lx or less
3-D data: 1/3-inch frame transfer CCD (340,000 pixels)
Color data: 3-D data is shared (color separation by rotary filter).
3-D data: 640 x 480 (for FINE mode); 320 x 240 (for FAST
mode)
Color data: 640 x 480
3-D data: Minolta format,
& (STL, DXF, OBJ, ASCII points, VRML)
(Converted to 3-D data by the Polygon Editing
Software/ standard accessory)
Color data: RGB 24-bit raster scan data
Compact Flash memory card
Total 3-D and color data capacity: 1.6MB per data
(for FAST mode), 3.6MB per data (for FINE mode)
5.7-inch LCD (320 x 240 pixels)
SCSI II (DMA synchronous transfer)
Commercial AC powerl00 to 240V (50 to 60Hz),
Rated current 0.6A (when 100Vac is input)
213 x 413 x 271 mm (8-3/8 x 16-1/4 x 10-11/16 in.)
Approx. 11 kg.
Temperature: 10-400C (50-1040F); relative humidity 65% or less
with no condensation, Pollution degree:2, Installation category:II
-10 to 500C (14-1220F); relative humidity 85% or less (at
350C/950F) with no condensation














CHAPTER 5
PLAXIS VERSION 7.2--FINITE ELEMENT CODE FOR SOIL AND ROCK
ANALYSIS

Introduction

Plaxis was developed in 1987 at the Technical University of Delft, initiated by the

Dutch Department of Public Works and Water Management. Initially, Plaxis was

developed to analyze river embankments in the soft soils of the Dutch lowlands.

Throughout the original development and later improvements, Plaxis extended its

applicability to cover most areas of geotechnical engineering. Plaxis excelled over the

years forming a company in 1993, Plaxis BV.

Plaxis is a computer program designed to provide a practical analysis tool for use

by engineers who are not specialists in finite element analysis. Non-linear finite element

computations done without a computer can be too time-consuming for regular analyses.

Plaxis 7.2 is an MS Windows-based program with easy-to-use tabs used to navigate

through the analysis.

Plaxis is a finite element program designed for the analysis of deformation and

stability in geotechnical engineering projects. Geotechnical engineering uses advanced

models for the simulation of non-linear and time dependent behavior of soils. Soil is a

multiphase material with properties that can change with a changing environment. Plaxis

is equipped to deal with hydrostatic and non-hydrostatic pore pressure in the soil. Even

though modeling of the soil itself is important, modeling of soil-structure interactions is










the application seen in many engineering projects today. Plaxis is also equipped with

features to analyze a number of aspects dealing with complex geotechnical structures.

Plaxis is made up of four internal programs, Input, Calculation, Output and Curves.

Graphical input of geometry models consists of soil layers, structures, construction

stages, loads and boundary conditions all created with drafting procedures on a CAD

(Computer Aided Drawing) screen, as shown in Figure 5.1. The CAD interface allows

accuracy and detailed modeling of real engineering situations. From the geometry

entered, a finite element mesh is generated. The automatic mesh generation feature

allows for fully automated mesh generation of unstructured finite element meshes.

w Lpd i'h
srcus. PMipe are moa dee. using te tr Ho












n as o s a 10o M D Im sm









P.rl e ; 1.7 uwl. ; Mailrm n

Figure 5.1: Plaxis 7.2 Computer Aided Drafting screen used to create modeling analysis.

Beam elements are used to model retaining walls, tunnel linings and other

structures. Pipes are modeled using the tunnel feature. Behavior of each element is

defined as flexural rigidity, normal stiffness and/or ultimate bending moment. Assigned









to the beam elements is a feature called "interface." Interfaces are joint elements used in

calculations of soil-structure interactions. For example, interfaces can simulate a zone of

intense shearing stresses located at the contact plane of footings, piles, geotextiles,

retaining walls and pipe surfaces. Values of friction angle and adhesion properties can be

assigned to interface elements. Anchors and geotextiles are additional features available

for modeling. The tunnel feature option creates circular and non circular tunnels

composed of arcs. Beams and interfaces may be added for analysis of tunnel linings or

interactions with the surrounding soil.

A number of soil models are used in Plaxis for analysis of soil performance. Mohr-

Coulomb, a simple non-linear model, is the most used model based on soil parameters

encountered in everyday practical situations. Other advanced soil models are also

available for analyses.

Pore pressures are analyzed within the Plaxis model. Steady state and excess pore

pressures are defined by the water table location in the model. There are two approaches

for steady state pore pressures. Complex pore pressure distributions are generated on the

basis of a two-dimensional groundwater flow analysis. For simple conditions, multi-

linear pore pressure distributions can be directly generated by a simple phreatic line in the

model. Excess pore pressures are computed during plastic calculations when pores are

full of water and subjected to loads.

A typical analysis with Plaxis finite element modeling involves input parameters

and specifications of model types. The input model is then run through the calculation

phase to produce output results for analysis. Plaxis is also equipped with a curves







56



program that produces graphs for analysis results, such as stress/strain curves and


load/displacement curves.


Plaxis--Input

The input program is the first step of a Plaxis analysis. At the start of a new project,


the General Settings window appears (as shown in Figure 5.2), prompting the user to set


the basic parameters of the project. The General Settings window has two tabs: Project


and Dimensions. The project tab allows the user to enter a description of the project, the


type of model used (i.e. plane strain) and the number of nodes to be used in the finite


element analysis (i.e., 6 and 15 node). The Dimensions Tab specifies units of length (ft),


force (lbs.) and time (day); and the geometry dimensions for the CAD screen and grid


spacing.







.m0 50D a0m 'M low0 1M0G 000 200
I_ II I I.N llI_ l ..I. I.g I _J,|.. III~ Lll ..l.... ='- ..J |. ... .gl..


l.ai E# r=...enu, |


praeer -



.10.00 i ..... I








PaIr rilblo ca-ara l
ure 5.2 s w1io in P i 7



M T. 1.3 m. *' MUCK O C
Figure 5.l General Settings window in Plaxis 7.2.
Figure 5.2: General Settings window in Plaxis 7.2.









Once the general setting parameters are entered into Plaxis, the CAD screen

appears for geometric creation of the model. An example of a tunnel model, as used in

this research, will be used to demonstrate a typical run-through of a Plaxis 7.2 finite

element modeling. The tunnel model can be used to simulate buried pipes for analyses.

The model described in the example is a four-foot diameter reinforced concrete pipe,

located within a soil box. Ultimately, a steel framed box will be designed to test concrete

pipes. Plaxis modeling will be performed on the soil structure interaction and in

analyzing the stresses induced on the sidewalls and pipe, thus defining the design

parameter for the box.

After the general settings are complete, the geometry of the model is created in the

input drawing area using the points and lines feature. With the Windows based program

the creation of a model is accomplished by working from left to right with the icons at the

top of the screen, as shown in Figure 5.3. Points and lines are used to create an enclosed

box of dimensions 20' x 10' x 10'. The tunnel option is then used to create a circular

tunnel representing the four-foot diameter concrete pipe, composed of arcs defined by a

radius and a radial increment (angle). By clicking on the tunnel feature, the user is

prompted with a choice to use a whole tunnel or a half tunnel. After this choice is made,

a window is opened as shown in Figure 5.4 to allow input of the radius and radial

increment. In this same window an interface and lining material is assigned to the tunnel

model.

For the example, a radius of 4 feet is entered and a tunnel and interface lining is

assigned to the tunnel. An interface is used to model soil-structure interaction and the









lining material provides the option to assign concrete properties to the tunnel lining. The

next step in the model is load and boundary conditions.

R U Hi b jar uin Rtel Htf m w


\1 111 000L0 t ,i *144 MIMI
Figure 5.3: Plaxis 7.2 Main Toolbar.























Figure 5.4: Tunnel Designer in Plaxis 7.2.
Sa a a ii i m a aa iii a











traction loads (distributed loads). Both features are chosen from the standard toolbar as-

shown in Figure 5.3. The soil box will be modeled as a rigid box with minimal











deflection. Fixities are prescribed displacements equal to zero. These ixities can be
applied to points or lines. By clicking on the standard cities button on the toolbar, the


i l i -,i--- J .
--i-





.: *I. -- *



Figure 5.4: Tunnel Designer in Plaxis 7.2.

The load and boundary conditions used for this model are standard fixities and

traction loads (distributed loads). Both features are chosen from the standard toolbar as

shown in Figure 5.3. The soil box will be modeled as a rigid box with minimal

deflection. Fixities are prescribed displacements equal to zero. These fixities can be

applied to points or lines. By clicking on the standard fixities button on the toolbar, the

two sidewalls and the bottom of the box were assigned both horizontal and vertical

fixities as shown in Figure 5.5 below.










After the standard fixities are assigned, moving to the left on the toolbar, the loads

are assigned through the traction icon as shown in Figure 5.5. A distributed load was

assigned using traction A, assigning the load to the two top points of the box on the CAD

drawing area. Plaxis will ask the user to enter a multiplier for loading steps in the

calculation phase. The default is 1.00, representing a true value for the load entered in

the calculation phase. A multiplier of 1.00 was used in the example.


Ph Eh .il CMAr l nd a '4k. D;jd rr4ll



.Ida, i_ ,. r,. '"
.. I III


rai-iir- ia








--f
4P ..





F" a.- ", I
:,a ~.rq ..m :IL.xCR I Ii

Figure 5.5: Standard Fixities in Plaxis 7.2 shown on a soil box with right half tunnel.

Once geometric definition of the model is finished, with all beam elements defined,

fixities assigned and loads designated, the material properties need to specified and added

to the model. Material data sets for soil, interfaces and beams are entered using the

material data set icon located on the main toolbar. When the icon is clicked, a Material

Sets window will open with choices for the project database as shown in Figure 5.6.









The example requires material data sets for soil & interfaces and beams. The first

step is to set the parameters for the soil & interfaces. Input parameters for the soil are

entered in the window shown in Figure 5.7 and include material model, unit weight,

permeability, stiffness, poisson's ratio, cohesion, friction and dilatancy angles and

interface properties. The input parameters used for the example soil model are as

follows:

Model used: Mohr Coulomb
Material type: drained
Unit weight: 120 pcf
Permeability: 0 (both vertical and horizontal)
Stiffness (E): 489,600 psf
Poisson's ratio (v): 0.3
Cohesion (c): 1 x 105
Phi angle ((p): 35.000
Dilatancy angle (x): 5.000

Specified along with the input parameters for the soil model are the settings for the

interface. The soil interface is specified as either rigid or set manually. This feature is

used for soil-structure modeling. The example model shown is modeling the soil-

structure interaction between the soil and a concrete pipe. Both rigid and manual settings

for the interface are used. Rigid assigns the soil properties to the interface and Manual

allows the user to specify a friction angle for modeling two materials in contact with one

another, such as concrete and soil or steel and soil.







61




71m1.i 1MlaI-L,-JF9 i

\Bi 1-- (7 3 I












a r+1 Ie M QJ







Figure 5.6: Material Sets Window in Plaxis 7.2.

=f.. .. ,








Er.
+M.1 I- .,
"I Ian -























.I. I I ...NJ 0 "- I i








Figure 5.7: Soil Input in Plaxis 7.2 Mohr Coulomb Model.

The material set for beams is specified when the user selects beams from the


material set window and assigns the parameters. Input parameters are entered in the










window as shown in Figure 5.8. The input parameters used for the beam material model

representing the concrete pipe are:

Axial Stiffness (EA): 107.3 x 106 lb/ft
Flexural Rigidity (El): 193.5 x 106 lb-ft/ft
Unit weight: 150 pcf
Poisson's ratio: 0.15

The axial stiffness and the flexural rigidity are calculated using the modulus of

elasticity and the cross sectional area of the concrete pipe along with the moment of

inertia of the pipe. An estimate for the unit weight was obtained from a concrete manual.


M -1+ .- -M i --..




.. i .i i . i ai.
._ ... h^






Si i i i i --


IJ m~ 1 j
I d I T *



Figure 5.8: Beam Properties Input Window in Plaxis 7.2.

After the material parameters are assigned to the model, the material sets window

appears once more to click and drag the material properties onto the model. Once this

task is complete the window is closed. At this point, the model setup is complete.

In order for the finite element analysis to proceed, however, the geometry must be

divided into elements called a finite element mesh. The mesh generation icon is the last

icon on the main tool bar, located to the far right. Plaxis is equipped with a self-









generating mesh tool. Once the mesh is generated, a plot is displayed through the output

program. The mesh can be defined as very coarse, coarse, medium, fine and very fine. A

finer mesh is used where large stress contributions might be seen due to loading. After

the mesh is generated the initial conditions are entered as shown in Figure 5.9.

Initial conditions involve an initial stress state before loading and an initial

situation. This process is still part of the input program. The initial conditions consist of

two different modes; one for the generation of initial water pressures (water conditions

mode) and one mode for the specification of the initial geometry configuration and the

generation of an initial effective stress field (geometry configuration mode). Switching

between these two modes is done by clicking on icons as shown in Figure 5.9.


7--1 ,i i ,
HI I -! I-lfi f I-- -i











Figure 5.9: Pore Water Pressure & Initial Stress Modes in Plaxis 7.2.

Water conditions are specified by means of the water weight and phreatic lines.

The model used for this example was in a dry state, thus there were no phreatic lines or

water pressures to generate.

The screen is then switched to initial geometry configuration. In this step the

example model is shown with the soil box and concrete pipe inside the box. The initial

geometry configuration allows the user to select geometry objects that are not active in









the initial situation. This means that the beam elements can be turned off to determine

the initial stresses before addition of the concrete pipe to the soil box.

The initial stresses are calculated using the Ko procedure with a default value of

0.426 when the stress icon is chosen on the tool bar. An initial vertical stress is

calculated using the coefficient of lateral earth pressure Ko. After the initial stresses are

generated, a window is displayed with the initial stresses showing the plane of direction.

Once the input stage is completed, the next step is the calculation stage.

Plaxis--Calculations

After the mesh is generated, the finite element calculations can be specified and

executed in the calculations program. In this program each type of calculation to be

performed is defined, along with the type of loading activated during the calculation.

Initiating the calculation program will open a window as shown in Figure 5.10. The three

tabs (general, parameters and multipliers) will allow the user to navigate through to the

end calculate command. The only settings to define in the calculation program are the

calculation type, selecting the calculation phases and multipliers. The other input areas in

the calculations program were set to default and were used as default for the example

model.























Figure 5.10: Plaxis 7.2 Calculations Program.
pti I l l l


















bottom of the calculations window. The initial phase represents the starting conditions of

the project as defined in the initial conditions mode of the Input program. For the
example model, two phases were added: staged construction and total multipliers. The

staged construction represents the installation of the concrete pipe with loading coming

from overburden soil above the pipe. The total multipliers stage is where the 1 assigned
Px 2 I Mtr *csfrqe Z V



Figure 5.10: Plaxis 7.2 Calculations Program.

In the calculation program under the general tab the calculation type is specified

along with the setup of the calculation phases) using the insert button as seen in Figure










distribu5.10. The insert button willvated add additional phases for calculation. In the example model

three phases were etup; initial phtab (as shown in Figure 5.11) the default modes are used.

The parameters and multipiis hers when highlighted in theruction phase display defined. The loading inputat the

bottom of the calculations which type of loadinitia phase represents the starticular calculonditions of
the project as defined in the initial conditions mode of the Input program. For the

example model, two phases were added: staged construction and total multipliers. The

staged construction represents the installation of the concrete pipe with loading coming

from overburden soil above the pipe. The total multipliers stage is where the assigned

distributed load is activated and applied to the model.

Under the parameters tab (as shown in Figure 5.11) the default modes are used.

The parameters tab is where the staged construction phase is defined. The loading input

group is used to specify which type of loading is considered in a particular calculation







66


phase. Loading input was set to stage construction. The define button, located at the

bottom right, will open the input window of the model allowing the user to deactivate and

activate soil clusters and structural objects that define the construction of the model.

After the staged construction phase is defined, the update key will return the user to the

calculation program.

The multipliers tab is where loads are assigned for the total multipliers stage.

There are two types of multipliers: incremental and total as shown in Figure 5.12. In the

example model shown, the total multiplier assigned in the input program is IMloadA at a

value of 16,000 with units of pounds per square foot. EMloadA controls the magnitude

of the traction loads as defined in the Load System A of the input program.



St I.' l .. i i- ...*. l
D tf a jUT II f a.. i ... .......... g i li a I+ ._a .l,.
--





S -- ,- a r




Fy~Z-





Fiur 5 ,. 7. lcai



Figure 5.11: Plaxis 7.2 Calculations Program Parameters Tab







67


r .. i. I V
L- l i' .'3 n* *. "l f .'- n ., -3 ... .. ..- ... i g I s .ml i. .. .




a -0.& a 7 E *..E -,

jn


I-. I


Figure 5.12: Plaxis Calculations Program Multipliers Tab

Another feature in the calculation phase sets up points for curves generated in the

curves program. The points can be entered by selecting the Select Points For Curves

option from the View menu or by clicking on the corresponding button in the tool bar.

Selection occurs when the output program opens showing a plot of the finite element

mesh displaying all of the nodes. Nodes are selected by left clicking the mouse on the

node of interest. Each node selected is characterized in the curves program by an

alphabetical letter. A node can also be deselected by clicking on it again with the mouse.

In the example model shown, points were selected to create curves for load

displacement and stress/strain curves. After selecting the points for curves the calculate

button will run the calculation program. A window is opened to view the loading

increments displaying different properties. Once the calculation is completed the

calculation window appears with green checks beside the phases of calculation. Output

of the calculation is viewed with the output program, which is opened by clicking on the

output button located at the top of the calculation window.









Plaxis--Output

The main output quantities of a finite element calculation are the displacements at

the nodes and the stresses at the stress points. When a finite element model involves

structural elements, structural forces are calculated in these elements. The output

program is equipped to display the results of a finite element analysis. The resulting

program window is displayed in Figure 5.13, showing the deformed finite element mesh

due to the load.

Results are viewed via the output program, displaying deformations, deformed

mesh, total displacements, total increments, total strains, incremental strains, stresses,

effective stresses, total stresses, plastic points, active pore pressures, excess pore

pressures, groundwater head, flow field, structures and interfaces, beams, geotextiles,

interfaces and anchors. In the example model (a four foot diameter concrete pipe

modeled in a 20' x 10' x 10' soil box) the only outputs of interest are deformed mesh,

total displacements (how much the pipe and soil settled), stresses, structures and

interfaces and beams (concrete pipe). Each of the output results can be viewed as a

picture, table or curve.

The first plot the user views in the output program is the deformed mesh plot as

seen in Figure 5.13. From this plot the user can navigate to other results by selecting

from the menu at the top of the output window. In the example model, the first thing to

analyze is the stress distribution throughout the soil contained in the box. Figure 5.14

displays the stress distribution throughout the soil box by means of shading, with red

indicating the highest stresses. Shading provides a colorful way of presenting results for

effective mean stresses. Another way to display stress distribution results is through the

contour plot of effective mean stresses as shown in Figure 5.15. The contour plot









69




displays a legend on the right side of the window that identifies each contour. A scan



line, labels the contours with respect to the legend.



gimHia!^ Eiq^Bi --t


-mvi m "o m am r m m m '..
I. i. .... I.I.. I I .. ..... I ... .


1._ is s a nar7 im -. i \


t J, -4 | ,j |_. .




S5.




F e a 3 d--


displayed on example model.


1% 4.- 4 ., mC.b |I --


-3


&ib,,

faxtla
-j.ton







rima,
amma





r1(mmW

LRn ax


RI A K 411 P'!SLIDw- mi IW Ol lt% .U

Figure 5.14: Plaxis Output Effective Mean Stresses Displayed by Mean Shading.


...... ......... .......


. I . n . I











111 1 M lal~il I Id|B| I-IIM

SF. ,- .. L 'si.. ...












i -
[ i)
,rI )" ,7


o I IOD
I 'WOM

r, aec

an" d


Asrl thnu
J 0-1ni Il- %--i 40%l4w
Figure 5.15: Plaxis Output Effective Mean Stresses Displayed by Contours.

Cross sections through the model can be used to present results of displacement and

stress distributions. Viewing output in a cross section allows the user to gain insight into

the distribution of a certain quantities calculated by the model. Cross sections are created

in the model by selecting the cross section button and then clicking and dragging a line

through the model where desired. In the example model stress distribution on the

sidewalls was desired, requiring a plot of stress distribution along the sidewalls of the soil

box. The cross section at the sidewall is shown in Figure 5.16. Also shown with the

cross section tool are the horizontal displacements along the cross section shown in

Figure 5.17.

Beam elements, as created by the input program, are viewed in the output program

with element and interface results. Element results are the beam forces, displacements

and bending moments. Interfaces assigned to the element will show displacements and

stresses acting at or within that interface. Displacement output for the beam element -


~;;;;,


. ID .


i







71


concrete pipe and the interface assigned to the concrete pipe are shown in Figure 5.18.

Plaxis will specify the extreme displacement for the beam elements and the interfaces.


The bending moment of the concrete pipe in the example model is shown in Figure 5.19.

Beam element properties are displayed in the output program by double clicking on the

element itself. The various display choices are activated from the sub menu found at the

top of the screen.





S is ..e.. I ..




i ,. a_ i I i_ lull










a B-
Figure 5.16: Stress Distribution Cross Section A-A in Plaxis 7.2 Output Program
-_






I i I il lqa~.w i i~ f I1 ,Iu .a .l..
Figure 5.16: Stress Distribution Cross Section A-A in Plaxis 7.2 Output Program














-1'.'


Fh i A SML n Ml n l
fl~jm~~~j~jr Fb ~ T


J.1

Figure 5.17:


I a lil w m i 1


a


i. lll 4* I a *a 1
Fn alli aSdisl


Horizontal Displacement Cross Section A-A in Plaxis 7.2
Horizontal Displacement Cross Section A-A in Plaxis 7.2


Output Program


I II a O a, ..lt .l'AD


... I.. ... I.l....I I.... I.. I i 1, I |...I I. ,


Ei m.pah'. r.eil 1*11
lCLD aU


ES ,fLs .t .. 31.
I~~uii~ U..smll


Figure 5.18: Displacements for the Pipe and Interface Plaxis 7.2 Output Program.


nmmr rnmrnr Ec...


~rriiiii
~o a~
I


.II





''
hL 54~,:












>JgJ *j
ajflJl


?a't ts i-n~a.~ P ~ns t.




















,,


A-fi ie jtff w *


Figure 5.19: Bending Moment for the Pipe Plaxis 7.2 Output Program.


The output program is also capable of producing tabular data for analysis. The


numerical data can be viewed in output tables for all types of graphical output by clicking


on the Table button in the main tool bar or by selecting the Table option from the View


menu. A menu is available to view selections of other quantities. Tables available in


Plaxis include displacement, stresses and strains, and stresses and forces. In the research


example used, tabular data was not used directly. Output data was fed into the curves


program to allow generation of different graphs.


~rmm umn llln~ll r-~-~5~--r~I~l~r .....;;.........


~D m
...1...~1....1.~. .1....1....1....1.. ~.


~-














CHAPTER 6
FINITE ELEMENT ANALYSIS OF A SOIL BOX TEST FACILITY

Introduction

A finite element analysis was performed in order to provide the necessary

information for the design and fabrication of a soil box for testing two different concrete

pipes, fiber reinforced and standard reinforced. Two and three-dimensional finite

element analyses were done in order to evaluate the stress on the sidewalls due to the

applied load and the effect of boundary conditions (i.e., friction on the sidewalls and

around the perimeter of the test pipe).

The soil box will be fabricated of steel and filled with compacted soil similar to that

used by the State of Florida in highway right-of-ways. The two-dimensional program

used, Plaxis 2D, is a finite element analysis for soil and rock. Plaxis 2D evaluated the

stresses on the sidewalls and the soil structure interaction between the soil and concrete

pipes as well as between the soil and steel structure sidewalls. From the two dimensional

analysis, the box dimensions were examined and the stresses evaluated using an interface

between the soil and concrete pipe and the soil and the sidewalls. In the analysis,

boundary conditions were evaluated in order to determine size requirements that would

not induce lateral stresses on the test pipe. The results from the two-dimensional analysis

are included in Appendix A.

A three-dimensional finite element analysis, Plaxis 3D Tunnel was used to evaluate

the two-dimensional analysis, plus stresses in the z direction, the third dimension. Plaxis

3D analyzed the stress distribution throughout the box while varying the friction angle at









the soil-sidewall and soil-pipe interfaces to examine the stress concentrations induced

within the soil box. Each of the three possible lengths to be considered (10', 15', and

20') was evaluated while varying the friction angle and analyzing the boundary

conditions. Displacements of the concrete pipes were evaluated with and without friction

on the sidewalls. Stresses were examined along the length of the pipe to ensure that no

shear stresses were induced on the ends of the pipes near the front and rear walls.

Two different manufactured pipes will be tested inside the soil box, fiber reinforced

and standard reinforced concrete pipes. The proposed pipe diameters to be used for

testing are 18" and 24" with the further possibility of sizes as large as 48" in diameter. In

order to ensure proper bedding, a depth of at least one diameter below the pipe inside the

soil box will be used in the analysis. For example, the 24" diameter pipe will need two

feet of underlying soil, resulting in an approximate height of six feet above the crown of

the pipe to the top of the box. An overburden soil depth of six feet will allow a

distribution of stresses that simulates in-situ conditions. The maximum load applied on

the soil will be 16,000 lbs/ft2.

The two types of backfill used in the analysis (shown in Table 6.1) were loose and

dense compacted soil. For the loose compacted soil, a Young's Modulus of 216,000

lbs/ft2 was used, while the dense compacted soil had a Young's Modulus of 489,600

lbs/ft2.

Input parameters for the different sizes of SRCP and FRCP are displayed in Tables

6.2-6.4. Young's Modulus for the FRCP and SRCP were obtained from previously

published literature. The SRCP modulus value was calculated using the American

Concrete Institute (ACI 318) relationship between elastic modulus and compressive









strength, as shown in Equation 6.1. Compressive strength for concrete pipe normally

ranges from 4000 lbs/in2 to 6000 lbs/in2 (Rinker Materials, 2003). Elastic modulus for

the SRCP was calculated using a compressive strength (f') of 4000 lbs/in2. The elastic

modulus for the FRCP, used to determine normal stiffness and flexural rigidity was 25

Gpa or 3.62 x 106 lbs/in2.


Ec = 57,000 f


(6.1)


Table 6.1 Material Properties of the soil (Loose & Dense).
Parameter Name Loose Dense Unit
Material Model Model Mohr-Coulomb Mohr-Coulomb
Type of material behavior Type Drained Drained
Soil weight above phr. level Yunsat 120 120 lbs/ft3
Soil weight below phr. level Ysat 120 120 lbs/ft3
Young's modulus Eref 216,000 489,600 lbs/ft3
Poisson's ratio v 0.3 0.3
Cohesion Cref 0.00001 0.00001 lbs/ft3
Friction angle (P 35 35 DEG.
Dilatancy angle y 5 5 DEG.


Table 6.2 Material Properties of the 18" diameter concrete pipes (FRCP & SRCP).
Parameter Name FRCP SRCP Unit
Type of behavior Material type Elastic Elastic
Normal stiffness EA 147,372,845 280,303,765 lbs/ft
Flexural Rigidity El 38,138,787 66791264 lbs-ft2/ft
Weight W 150 150 lbs/ft2
Poisson's ratio v 0.15 0.15


Table 6.3 Material Properties of the 24" diameter concrete pipes (FRCP & SRCP).
Parameter Name FRCP SRCP Unit
Type of behavior Material type Elastic Elastic
Normal stiffness EA 198,636,143 382,227,687 lbs/ft
Flexural Rigidity El 93,303,588 168,720,067 lbs-ft2/ft
Weight W 150 150 lbs/ft2
Poisson's ratio v 0.15 0.15









Table 6.4 Material Properties of the 48" diameter concrete pipes (FRCP & SRCP).
Parameter Name FRCP SRCP Unit
Type of behavior Material type Elastic Elastic
Normal stiffness EA 665,723,216 1,228,267,715 lbs/ft
Flexural Rigidity El 1,263,829,838 2,322,073,435 lbs-ft2/ft
Weight W 150 150 lbs/ft2
Poisson's ratio v 0.15 0.15


Plaxis 3D--Verification Analysis

A three-dimensional analysis was performed to verify the results obtained from the

two-dimensional analysis, as well as to analyze the stresses and displacements in the third

dimension. Plaxis 3D introduces a third dimension of analysis allowing stresses in the z-

direction to be examined. From the two-dimensional analysis, results showed that the use

of well-compacted soil (90% proctor) provided proper stability with half as much

settlement as that of the loose compacted soil. It is important to obtain proper

compaction during the installation of the pipe to produce maximum performance of the

pipe.

Three-dimensional finite element analysis was used to verify eight analyses

performed with the two-dimensional finite element program (Appendix A). Verification

of the stresses and displacements throughout the soil box were done for the 18" and the

24" diameter FRCP and SRCP. The ability to test a 48" diameter pipe in the soil box is

questionable according to the two-dimensional analysis due to the limited depth of soil

cover on top of the test pipe.

Verification of the 18" and 24" diameter pipe was done using the three dimensional

version of Plaxis. Four different analyses of the 18" and 24" pipe were performed using

dense compacted soil and varying the friction angle along the wall within the interface for

each of the SRCP and FRCP. The three-dimensional analyses differed very little from









the two-dimensional analyses, through the three-dimensional results provided better

stress distribution throughout the soil box in the third dimension. Varying the soil-wall

friction angle from a value equal to the soil friction angle of 35 degrees to less than 5

degrees doubled the amount of total displacements throughout the whole soil box.

Assigning a value of 5 degrees to the interface resulted in higher displacements within the

soil upon loading.

Friction on the sidewalls equal to that of the soil resulted in a total displacement of

0.236' while a friction angle of less than 5 degrees on the sidewalls results in a total

displacement of 0.433'. When comparing the 18" and 24" diameter pipes, the total

displacements were the same. Figure 6.1 shows the total displacements for the 24" FRCP

and Figure 6.2 shows the total displacement for the 24" SRCP, both with a friction angle

equal to that of the soil. It is visually apparent on the sidewalls that there is friction. The

load is distributed throughout the third dimension and the friction along the wall is visible

through the shading.

The effective mean stress reported from the three-dimensional analysis differed

little from the two-dimensional analysis. For example, the extreme effective mean stress

for the 18" SRCP, friction sidewalls, measured 12,690 lbs/ft2 in the two-dimensional and

12,520 lbs/ft2 in the three-dimensional analysis. The three-dimensional analysis

distributes the stresses more efficiently in the z direction (i.e., the third dimension).

The analysis of the box focuses on the maximum stress on the sidewalls as a result

of the 16,000-lbs/ft2 load. Each of the two-dimensional wall friction analyses examined

the effective mean stress on the sidewalls. When verified by the three-dimensional finite












































Figure 6.1 A Three Dimensional View of Total Displacements for 24" FRCP.


Figure 6.2 Three Dimensional View of Total Displacements for 24" SRCP


* CdS ~ ~ hlbY1 ~~Ltm r"L *n -"u
!e a p P q d ~g ;- -I
,cei~-lcrraiCII Clll(b*rriH-~i








80



element program, it was found that the sidewall stresses were slightly smaller. Sidewall


stresses examined through the interface along the sidewall increased when the friction


angle was set to 35 degrees.


For example, the interface wall stress for the 18" diameter SRCP with friction was


7,070-lbs/ft2 compared to an interface without friction of 4,470-lbs/ft2. The soil was


unable to move freely upon load application when friction occurred at the sidewalls,


therefore creating higher stresses on the sidewalls. Figure 6.3 shows the left sidewall of


the soil box displaying the effective normal stresses.


[Dbn/ft l

2250.000
-2500.000
-2750.000
-3000.000
-3250 000
-3500.000
-375 0000
-4000 000
4? n nnU



-5000000
-5250.000
-5500000
.iRn nnn



-6500 000
-6750 000
-7000.000
-7250 000


fller.li.e normal sllesnls

Figure 6.3 Left Side Interface of Soil Box Model 24" FRCP


An interface was also assigned to the perimeter of the pipe inside the soil box for


soil-structure interaction modeling. The friction angle associated with this interface was


also varied in the same manner as the sidewalls. The extreme effective mean stress






81


around the pipe and the extreme total displacement around the pipe are reported for each

of the tests. There was very little difference between the FRCP and the SRCP modeled.

In varying the friction angle at the interface around the perimeter of the pipe, the effective

mean stress increased greatly, from -2,860 lbs/ft2 for a friction angle of less than 5

degrees to -27,870 lbs/ft2 when the friction angle was set to the same value as the soil.

Results are depicted graphically in Figures 6.4 and 6.5.























Figure 6.4 18" Diameter FRCP Friction Area on Surface of Pipe.

In situ conditions reflect a soil interacting with a concrete surface and soil-soil

interaction around the pipe. In an attempt to model this, the friction analysis along the

sidewall is justified but the friction around the perimeter of the pipe is questionable. It is

impossible to have near zero friction around the perimeter of the pipe that would result in

a stress decrease, as seen in the three-dimensional analysis for a friction angle less than 5

degrees. FRCP is a smoother pipe, when compared to SRCP, yet the friction will not be






82
























Figure 6.5 18" Diameter FRCP Near Zero Frictionless Area on Surface of Pipe.

near zero. Thus using a friction angle equal to that of soil around the perimeter of the

pipe produced more justified results. More stress is induced as friction is encountered

throughout the soil depth under the load.

The displacement for the interface reacted in the same manner increasing, for

example, from 0.072' to 0.211' when the friction angle was increased to a value equal to

that of the soil friction angle, 35 degrees. This is a direct result of the soil encountering a

rough friction surface along the pipe. Table 6.5 presents the results of the three-

dimensional analysis verification of the two-dimensional analysis. Note that the negative

values refer to compression.

Other concerns that were addressed with three-dimensional finite element analysis

included stresses along the length of the pipe as a result of a possible shear stress induced

on the ends of the pipe due to wall friction and displacement of the entire pipe as a result

of both wall friction and service load.










Soil Box Analysis

A three-dimensional finite element analysis was conducted with and without the

pipe inside the soil box to analyze the box dimensions while varying the friction angle

through four different analyses. An approximate height often feet was used for the

height of the soil box based on providing proper bedding needs for each of the 18" and

24" diameter pipe. A bedding depth of one pipe diameter below the invert of the pipe

was used in the finite element analysis. Three different lengths of soil box were

analyzed, ten, fifteen and twenty feet. Full-scale models were used for analyses

eliminating any concerns of the results.

Table 6.5 Three Dimensional Analysis Verification ofPlaxis 2D Wall Friction Analysis


Pipe
FRCP
SRCP
FRCP
SRCP
FRCP
SRCP
FRCP
SRCP


D1 C2
18" Dense
18" Dense
18" Dense
18" Dense
24" Dense
24" Dense
24" Dense
24" Dense


R EPSs EMS6 Dispi IEWS7 IEPS8 Disp2 PM9


1
1
0.05
0.05
1
1
0.05
0.05


-23650
-21600
-24430
-24440
-24520
-22180
-26430
-26450


-13830
-12690
-15400
-15400
-14690
-13380
-14760
-14750


0.236
0.237
0.331
0.331
0.230
0.234
0.333
0.333


-7340
-7280
-4840
-4840
-7410
-7380
-4790
-4790


-24630
-22440
-4590
-4550
-24960
-22610
-5800
-5740


-0.060
-0.065
-0.143
-0.143
-0.696
-0.077
-0.194
-0.195


-6170
-5230
-961
-877
-11090
-9600
-2220
-2070


Disp3
0.057
0.064
0.090
0.090
0.067
0.076
-0.113
0.113


*All values reported are extreme values (i.e., the maximum values).
1. D is the pipe diameter in inches.
2. C is the compaction of the backfill.
3. R is the interface value of strength. A value of 1 represents a friction angle the same as the
backfill soil. A value of 0.05 represents a friction angle less than five degrees.
4. DM is the deformed finite element mesh after loading 16,000 Ibs/ft2
5. EPS is the effective principal stress for the entire model box in Ibs/ft2
6. EMS is the effective mean stress for the entire model box in Ibs/ft2
7. IEWS is the interface effective normal wall stress in Ibs/ft2
8. IEPS is the interface effective normal pipe stress in Ibs/ft2
9. PM is the pipe bending moment in Ibs-ft/ft
Disp, is the total displacements for the entire box in feet.
Disp2 is the total displacements for interface around the perimeter of the pipe in feet.
Disp3 is the total displacement of the concrete pipe in feet.









The first analysis group examined the three different lengths with soil backfill only

and no test pipe in the box. In this analysis, the friction angle was varied from 35 degrees

to less than 5 degrees. This friction angle was assigned to the sidewalls. The remaining

three analyses examined boundary effects with a 24" FRCP inserted in the soil box.

Again each of the three soil box lengths was analyzed while varying the friction angle at

the sidewalls and around the perimeter of the test pipe. Note that the decision to use the

FRCP is not due to preference; it is merely an example to model the boundary effects

since Young's modulus for both SRCP and FRCP are similar in magnitude.

Soil Box Analysis--No pipe

A three-dimensional analysis was done on three different lengths of the soil box to

determine the stress and displacement of the soil due to a maximum distributed load of

16,000 lbs/ft2. An interface was assigned to the sidewalls to all variation of the friction

angles. Dense compacted backfill soil was used since it provides better bedding for the

test pipe than the loose compacted soil. Six models, two for each length, were analyzed

with a friction value of 35 degrees and less than 5 degrees.

Figures 6.6-6.8 show the stress distribution for the effective mean stresses

throughout the entire soil box with sidewall friction angle 35 degrees. Figures 6.9-6.11

show the stress distribution for a sidewall friction of less than 5 degrees. For the soil box

model with sidewall friction, the extreme effective mean stresses decreased as the length

of the soil box increased. For the friction value of less than 5 degrees, the extreme

effective mean stresses increased as the length of the soil box increased. A higher

displacement resulted in higher load concentrations throughout the soil box, thus the

increase in stress with an increase in length. Another observation is the stress

concentration and distribution throughout the soil. In comparing the 10' soil box with











friction to the 15' soil box with friction, the stress concentration around the middle of the


box increases approximately 2,000 lbs/ft2. This is important in dimensioning the soil box


for the maximum load needed for proper testing results.




-2500000ooooo
-3000 000
3500.000
40 0000


5500 U0j









Extreme effective mean stress 03103 b/ft2
J -85000






















largest effective normal stress occurs in the 15' soil box at a value of 7,280 Ibs/ft2. The
Erene elfotve m .120iO -- '"ft 2
Figure 6s6 10' Length Soil Box: Effective Mean Stresses with Sidewall Friction:

t tresses were modeled along the sidewall. Figures 6.12-6.14 show the left sidewall


deformation plane with a friction angle of 35 degrees. The sidewall interface with the

largest effective normal stress occurs in the 15' soil box at a value of 7,280 lbs/ft2. The


10' soil box has a sidewall interface stress value of 6,620 lbs/ft2 and the 20' length a


value of 7,150 lbs/ft2. The 10' box contains only half the volume of soil as the 20' length


box, resulting in the lower stress value. Also, the surface area of the 10' box exposed to


the distributed load is half the area of the 20' box. Figures 6.15-6.17 show the left


sidewall deformation plane with a friction angle less than 5 degree. The mean effective















normal stress along the sidewall of the soil box with a friction angle of less than 5



degrees


EIective mean stresrse
Extiem efIectve mean stress -11 5410 3Absft 2

Figure 6.7 15' Length Soil Box: Effective Mean Stresses with Sidewall Friction.


Effective mean stresses
E reme effctve mearr st re -11.0810 3 Ib ls 2

Figure 6.8 20' Length Soil Box: Effective Mean Stresses with Sidewall Friction.


[hs/ft 1


100000

-2000 000

3000 000

40000000






-7t0000




-9000000

-1000000

11000001

12000 001


















[Ibs/ft1

-560R000

-6000000




7200000
_ .. ....... ..,'


Figure 6.9 10' Length Soi Box: Effective Mean Stresses with Frictionless Sidewalls.
Figure 6.9 10' Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls.


A A
-A *A


A A
A


I .. -- -^-- .2 AOE
'i -.s ,1 ,. ,-"'" \ -.' ,* .,- r

\ ,' \ -t /, i l I I'* ' *"
" _-_-_-_- i --. -.-/ '-- ....
'', f ,
', i ', i "'
/ I ,r /
~ L


I' I


Figure 6.10 15' Length Soil Box: Effective Mean Stresses with Frictionless Sidewalls.


r


16000 000
I-55000001









10500 000
- 1,11,,f




-8 -IDOO0 OQ:










-125o000:

-1300000f
-13500.00


A


Prt ~r--r-i