t 115%.“; . . V 44hu~by glamfinyfi . if“ no A. g 1.251....» t..53}.1..:..t alt 3. in}: .6». .3. 13.4...h T)... 7:52 Fri}... lrl..3r..3:l I) in. 3}: 13:1... .21 “5".” Ii: 3.54: o: 121 .23". 4.,er . 3: .. uraérnnifif Ln. .mgmgaxég i 3?... . . E . V . . _ . _ E3“? ,. , .7 . . f Llanmfi Michigan State University \. PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before gate due, MAY BE RECALLED with ea rller due date If req uested. \ DATE DUE DATE DUE ' DATE DUE ' A I . - 7 6/01 c:/ClRC/DateDue.p65-p. 15 MPOSITE FACTORS INFLUENCING STRENGTH LOSS OF GLASS FIBECRRggE BARS IN THE ALKALINE ENVIRONMENT OF CON By Gustavo D. Dominguez A THESIS Submitted to Michigan State University in partial fulfillment Of the requirements for the degree of MASTER OF SCIENCE Department of Civil and Environmental Engineering 2002 ABSTRACT FACTORS INFLUENCING STRENGTH LOSS OF GLASS FIBER COMPOSITE BARS IN THE ALKALINE ENVIRONMENT OF CONCRETE By GUStaVO D. Dominguez . 0 ° . “Crete Steel corrosion is one of the prinCIpal problems present in reinforced co 0 O O . Steel’ construction. Billions of dollars are Spent annually for rehabilitation and repair 0f . S due to reinforced concrete infrastructure systems such as bridges and coastal sti'ucmre 05.10n . . . 1' 0011 deterioration caused by corrOSion of steel. As a result, materials With bette . . (Ross 6‘06 resistance and Similar mechanical prOpertieS to steel are on demand' ‘3 . sfififlg , Reinforced Polymer (GFRP) composite bars are corrosion-proof) provide l“ when . . . \65 are lightweight, and have superior chemical and electromagnetic p rope“ compared to ordinary Steel reinforcement. GF RP bars however, degrade when CXPOSCd over long periods Of time to the highly alkaline environments of co t more e, de d . . . . . r f among other gra ation issues. This investiga 101‘ 0011365 on the effect of high alkal‘ 111i 1‘ and bond strength of several types of GFRP bars. Accelerated aging tech” 3’ on {6175176 9 . I. employed to CXplore the effects of GP RP characteristics (fiber volume 5:30 9Ues were (1'0 com osition, b ' sence of rotective coatin . ’ In . p ar diameter, and pre P g) on their alkali an“, . . . . . 1‘ 68' Test results indicated that alkalinity has less impact on bars With 10 18’ W . . . . e1- fibe’ V01 fractions and on bars treated with PFOtective coating; mixed resu Its e . . . e e Obtained . regard to matrix type and bar diameter comparisons. Per the Part' in l Chlar bOnd te specimens investigated in this project, accelerated aging did not Chan ge b St 0nd strength bUt altered the failure mechanism associated with bond failure of the SpeCime To my late best fiiend Guillermo José Pescador, iii ACKNOWLEDGMENTS First, and most importantly I would like to thank God for providing me with the blessing of the Opportunity to get involved in this fabulous teaming experience and conduct this research. Unbounded thanks to my parents who raised me up and nev er Stepped guiding me towards the correct path to follow, and to whom I owe the place hat I am today. Many thanks to my adViSOI', Dr. Parviz Soroushian, who enlightened me so much and guided me through hardships and complexities of this research. 1 am very gratefuI for his patience and time during this Whole journey. My sincere thanks to the members of my thesis committee, Dr. Neeraj Bush and Dr. Rigoberto Burguefm’ for their help and assistance. A bCXQ Countless thanks to Mr- Siavosh Ravanbaksh f tafic’c provided during the entire duration of this research. Other great ilidi - duals to“ «0“; v1 v . . cry grateful are Mr. LC. Brenton for prOVIdlng “Sistance in the lab oiamy’ Mohamed Elzafraney for capturing SEM images, and Ms Co ' n... 5...... who provided assistance in statistical analyseS- I W0111d also 1ike to extend particular Daniel Wendichansky, Dr. Juan 3- 3611131: and Dr. undergraduate professors who inspired and motivated me towards gr :Oaado adieu Special thanks to m in d d 11 y en 3 an CO eagues for pl'OVld good and bad tirnes during my research. (”Um-(:81! e Finally, I’d like to extend countless 3”“de towards the eNationaI SCie en Ce Foundation and DPD, Inc, for promoting the exploration of Comp PPlicati infrastructure and for sponsoring my research. 0’18 in iv TABLE OF CONTENTS LIST OF TABLES .................................................................................. v11 LIST OF FIGURES ................................................................................ vnl CHAPTER 1 .. INTRODUCTION ................................................................. 1 1.1 GENERAL .................................................................................... 15 1,2 SCOPE OF WORK ......................................................................... -5 1.3 RESEARCH OBJECTIVES .............................................................. - CHAPTER 2 — LITERATURE REVIEW ........................................................ . 6 2.1 GENERAL ................................................................................... - 6 2.2 FIBER TYPES .............................................................................. - 8 2.2.1 Glass Fibers ......................................................................... - 3 2.2.2 Other Fibers ........................................................................ 1 3 2.3 MATRD< PHASE .......................................................................... 1 5 2.3.1 Overview .......................................................................... 1 5 2.3.2 ThermosetMatrices..................... . ., . ......................... 16 2.3.3 Thermoplastic Matrices ............................................ 1 8 2.3.4 Timers ................ 19 2.4 COMPOSITE FABRICATION ................................................. 20 2.5 MECHANICAL AND THERMAL PROPERTIES OF COMOgfii-fi """""""" 2.5.1 TensileStrength andElastic Modulus S """""" 22 2.5.2 Shear Strength ................................................ - WW22 2.5.3 Creep .............................................................................. 24 2.5.4 Thermal Expansion ............................................................... 25 2.5.5 Bond Strength. . ................................................................. 26 2.6 MECHANISMS OF GFRP COMPOSITES DEGRADATION. . ...' .............. '27 2.6.1 Overview ............................................................ ' 28 2.5.2 Alkali Degradation ............................................................... 28 2.6.3 Degradation Due to Other Chemicals .......................................... 23 2.6.4 Moisture Effects ................................................................ 31 2.6.5 Thermal Resistance .............................................................. 31 2.6.6 Degradation Due to Ultraviolet Rays Resistance ........................... 33 ............. 45 CHAPTER 3 - EXPERIMENTAL PROGRAM .............................. .................. 36 3.1 SAMPLE PREPARATION AND CONDITIONING .............. 3-2 TESTINGPROCEDURES...”....................................... """""""""" 36 3.2.1 Tension Tests ................................................................... 39 3.2.2 Pullout (Bond) Tests ....................................................... i: CHAPTER 4 -EXPERIMENTAL RESULTS AND ANALYSES . . . ..44 4.1 TENSILE STRENGTH OF GFRP BARS .............................................. 44 4.2 STATISTICAL ANALYSIS OF TENSILE STRENGTH TEST RESULTS. . ....5 1 4.3 MICROSCOPIC ANALYSIS OF GFRP BARS ....................................... :72 4.4 BOND STRENGTH OF GFRP BARS ................................................. . CHAPTER 5 — CONCLUSIONS ................................................................. 66 APPENDICES APPENDIX A — RAW TENSION TEST DATA ............................................... 69 APPENDD< B —— RAW BOND TEST DATA ................................................... . 80 REFERENCES ...................................................................................... 82 vi Number 2.1 2.2 2.3 3.1 3.2 4.1 4.2 4.3 4.4 A.l A2 A3 A4 B. 1 B2 LIST OF TABLES Title Page Comparison of Mechanical and Physical PIOperties of Different 9 Types of Glass Fibers Tensile PrOperties Of FRP Composite Bars and Grade 60 Steel 24 Reinforcement Coefficient of Thermal Expansion of FRP Composite Bars, Steel 26 and Concrete Tensile Strength Test Program 36 Bond Test Program 37 Mean Tensile Strength Values (MPa) Prior to and Aficr Aging 44 Mean Percentage Drop in Tensile Strength After Aging 45 P-values Associated with Different Factors 52 P-values for Analysis of Combined Data 55 Tensile Strength Results of GFRP Bars from Manufacturer I 69 Tensile Strength Results Of GFRP Bars fiom MaHUfactUrer 11 Tensile Strength Results Of GFRP Bars from Manufacturer III 73 Tensile Strength Results of GFRP Bars from Manufacturer IV 76 Bond Strength Results of GFRP Bars from Manufacturer I 77 Bond Strength Results of GFRP Bars from Manufacturer II 8:0 vii Numb er 2.1 2.2 2.3 2.4 2.5 2.6 2.7 3.1 3.2 3.3 3.4 3.5 3.6 3.7 4.1 4.2 4.3 4.4 LIST OF FIGURES Title Page Glass Fiber Melting Process 10 Breakdown of Si-O—Si Bonds in Glass Fibers ll Polymerization of Thermoset Resin l7 Filament Winding Manufacturing Process 21 Pultrusion Manufacturing Process 22 Typical Tensile Stress-Strain Relationship of Pultruded GFRP 23 Bars and Grade 60 Reinforcing Steel Alkali Attack in Pultruded GFRP Bars 29 Aging Tank with Alkali Solution 33 GFRP Bar Tension Test Specimen 39 Tension Test Set-Up 40 Diagram of Bond Test Sample 41 Setup of Bond Test 313601133113 Prior to Pouring 0f Concrete Detail of Development Length and Debonded Ends of Bar 41 Bond Test Setup 42 43 Mean Values and Standard Errors 0f GFRP Bar Tensile Strongth Test Results 47 Percentage DIOp in Tensile Strength (Mean Values and Standard Errors) of GFRP Bars After Aging 49 Least Square Means and Standard Error of % Drop in Tensile 52 Strength vs. Aging Duration . Least Square Means and Standard Error of % Drop in Tensile 53 Strength vs. Aging Condition viii 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 Least Square Means and Standard Error of °/o Drop in Tensile 54 Strength vs. Bar Coating Condition Least Square Means of % Drop in Tensile Strength vs. Fiber 56 Volume Fraction for all GFRP Bars Tested Afier Aging Least Square Means of % Drop in Tensile Strength vs. Bar 56 Coating Condition for all GFRP Bars Tested After Aging Scanning Electron Microscope (SEM) Images of Glass Fibers in 58 Bars from Manufacturer 1 (Vinylester Resin, Vf = 70%) at 350x Magnification Scanning Electron Microsc0pe (SEM) Images of Glass Fibers in 59 Bars from Manufacturer II (V inylester Resin, Vf = 60%) at 350x Magnification Scanning Electron Micros00pe (SEM) Images of Glass Fibers in 60 Bars from Manufacturer 111 (Polyester Resin, Vf = 60% ) at 350X Magnification Scanning Electron Microscope (SEM) 11113365 Of Glass Fibers in 61 Bars from Manufacturer IV (Polyester Resin, Vr = 50%) at SOOX Magnification Bond Strength Test Results for Uncoated and Coated in Unaged and Aged Condition GFRP Bars 62 Bond Strength Values for Different Failure Mechanisms 64 Bond Strength Failure Mechanisms 65 ix CHAPTER 1 INTRODUCTION 1-1 m Numerous research studies have concluded that exposure to high alkali environments have a detrimental effect On glass fiber polymer composites. This occurs because of the ability of the basic components in a highly alkaline environment to penetrate and attack the components of glass fiber. Glass fiber, being the main load— bearing component in a composite, degrades by this chemical process and thus, the composite as whole loses its strength and load-bearing capacity, on the other hand, glass fiber composites are unaffected by the effects of corrosion. For this reason, me use of glass fiber reinforced polymer (GFRP) rods is an attractive alternative to regular steel reinforcement in special applications such as structural components or Systems that are exposed to highly corrosrve envrronments. GFRP rods are also 11' ght‘veigbt, ”on- magnetic, and have low thermal and electric conductivity. Hydrated concrete possesses a PH ranging from 13'4'135’ making 1111' 8 Wide] used construction material highly alkaline. This alkalinity is a product of the p of Na+, K”: and Ca2+ ions in the water trapped in pores within concrete, also known as Pore water solution. When GFRP rods are used as concrete reinforcement the pore ’ Water solution penetrates the outer coating or layer of the resin of bars , reaches the fibers and deteriorates their strength capacity over time. The degree of strength loss depends on th e . 6 ° 0 ' ' . time of xposure to alkaline s lutron, temperature of the solution, paras”), of concrete, type of fiber used to manufacture the composite, composition of the matrix in composite, and the coating layer used around the bar (among other less influential factors). GFRP bars are typically manufaCtured through a process known as pultrusion in which, continuous fibers are run through a thermoset resin bath. The wet fibers are then passed through a series of dies that apply heat for curing of the matrix and give the bars their final size and shape. Pultrusion enables profiles of constant cross-section to be manufactured continuously. This requires that the constant fiber distribution and cross— sectional shape; no bends or tapers can be introduced. The process is analogous to extrusion of plastics and nonferrous metals. Different means of irriproving the durability characteristics of GFRP bars in the alkaline environment 01' concrete have been deve10ped. These include the change from polyester to vinerSter resin and introdUCllOIl of a resin-rich surface layer (With discrete fiber reinforcement). Even though thGSe measures have yielded some gains in alkali resistance of GFRP bars, “fiscal-eh h . , as shown that GFRP bars still deteriorate over short periods of accelerated aging in a highly alkaline environment. 1'9 Fibers constitute critical elements of composites; they are the main 1 oad~be o A . . arm component of the composrtes. Fibers generally occupy 30% - 70% of the 8 . 1n composites. They can be contmuouS, Chopped, woven, stitched, and/0r braided F'b . I ers are usuall treated with sizin s SUCh as St . . . y g arch, gelatin, 011 or wax to improve bond to binders and also to improve handling attributes.10 The fibers used to manufacture GFRP b ars tested in this research are continuous longitudinal E-glass fibers. Carbon and 'd fibers and other types of glass fibers have also been used to manufaCture composite rebar, but their high price makes their use in concrete reinforcement bars very limited. ’0 The primary role of the matrix in Composites is to bind the fibers together and protect them from the environment. Two types of resins exist: thermoplastic and therrnoset. Thermoplastics melt when heated and solidify when cooled. These long-chain polymers do not chemically cross-link because they do not cure permanently. Thermoset resins are made up by two components, namely a polymeric resin and a hardener. When both are combined at a specific proportion and cured by application of heat, tile thermosetting resin will cure permanently by irreversible Cross-linking at elevated temperatures. This characteristic favors structural applicati011$ of thermOSet resin composites. The most common resins used in composites are the unsaturated polyesters, epoxies, and Vinyl esters; Polyurethanes and phenolics are us ed less commonly. Thermoset resins exhibit better resistance to environment and desirable m b . properties.10 Epoxy resins are also used to manufacture comPOSite bars, but are cc 311mg] no . . t common due to their higher cost. melester has higher Physical Very ropel'ties tb POIyesterS, and costs less than epoxies. Bars tested in this study were manuf an actured . . . Us FOIYCster and Vinylester resms. 111g Fillers are added to the resin matrix for controlling material c ost an ' - . d Improvmg its meohanical and chemical properties. some composites that are rich in resins can be subj ect to high shrinkage and creep and may also provide low tensile Strength The three major types of fillers used in the composite £1]de are calcium carbonate, (kaolin, and alumina trihydrate. Other common fillers include mica, feldspar, wollastonite, silica, talc, and glasses. When one or more fillers are added to a properly formulated composite system, gains in performance include improved fire and chemical resistance, high mechanical strength, and reduced Shrinkage. Other improvements include increased toughness and fatigue life, and reduced creep, Some fillers cause composites to have lower thermal expansion and exotherm coefficient.lo Additives are used in the composites to improve the performance characteristics , aesthetics, and the manufacturing process. They are incorporated into the matrix system during mixing of the components. Additives can be divided into three groupSI (1) catalysts, promoters, and inhibitors; (2) coloring dyes; and (3) releasing agents ,‘0 In composite bars, a surface layer of distinct attribut es may be applied f 01' improved protection against the environment and for enhanced bond Char acten'sn'cs Tbi ‘ S layer may have the same base reSlIl as the matrix. Discrete fibers may b e 1. “Clad . reSin t 0 . .. . w [n the 0 improve cracking resistance of this surface layer. A coarse material Sac}; as Sand employed for this purpose are roving of strands of glass fiber around the laye cInes n 3 may also be used to imprOVe bond between the bar and concrete 0th 01’ press molding to create protrusions or irregular surfaces on the bar. 1.2 SCOPE OF WORK In this research, we produced comprehensive. data on performance of four commercially available GFRP bars (two of which are produced particularly for concrete reinforcement), We also conducted a thorough investigation of the effect of epoxy coating on tensile and bond stre'ltlg’th Of GFRP bars under accelerated aging conditions. A preliminary assessment was made 0f the effect of epoxy coating on bond of GFRP bars With Surface deformations to concrete' An epoxy Coating With 15% silica fume content (by Weight) was used; comparisons were made between the aging behavior of uncoated and coated GFRP bars in order to assess the effectiveness 0f coatings in protecting GFRP bars against attack by the highly alkaline enViI‘Orlment of concrete, For the purpose of this study, we focused on the mOSt economical and widely used CE-glaSS) fiber reinforced composite bars. RESEARCH OBJECTIVES 1.3 The focus of this study was to investigate the effect of the high alk . . alumy Of Concrete environment on tensile and bond strength of Glass Fiber Reinforced Pol Wher- Composite Bars. Also included in the scope of this study was to investigate the effects of fiber volume fraction, bar diameter, matrix type and epoxy coating on alkali 1’eSiStance of GFRP bars in concrete. CHAPTER 2 LITERATURE REVIEW 2. 1 GENERAL Reinforced concrete is the most “Addy “33d material for construction Of . ' d bui ' lflfiaStI'UCture systems such as bndges an ldmg Structures. It is well known mat . tivel - . concrete is strong in compreSSIOna but rela y Weak "1 tel18ion; a reinforcing materl a1 should thus be used to withstand “‘6 ““5116 Stresses in concrete. Steel has been the , rete r ’ - matenal of choice for use as °°nc emforcemem SInee the late 1800’s. Some Properties that have helped this long-lasung he between concrete and steel are the high tensile strength and low prime °f “661’ and the °°mpatibility in thermal expansion coefficients between both materials. The use of steel along With concrete in construction has 1101; b 6611 free 0 f problems. Steel corrodes in the presence of moisture and air, creating an eXp , anSIVe reaction that causes spalling and breakage of concrete and reduced steel Properties Thi ' S Phenomenon is pronounced in the presence of salt, and is evident in coastal Structure 8 Such as Seawalls and bridges, and also the case of bridge decks and parking Structures exposed to de-icing salts. Several methods have been used to prevailt steel COTTOSiOn from occurring, such as use of stainless steel instead of regular steel, epoxy coating and . - - 1 cathodic protection, but none of them prov1de a final or practical solution. The search for a new reinforcement material that provides good tensil tr e S ength and corrosion resistance has led to experimentation with Fiber Rainforced Polyme r materials commonly known as advanced composites Advanc . (FRP) . ed composrtes are made up of two main materials, namely a resin (matrix) and fibers. Fibers are the main load- bearing components while the matrix is in charge of binding the fibers together and protecting them fi'om the environment. The origins of these types of materials date back to post-World War 11 years, where the aerospace and defense applications prompted major development efforts concerning composites.ll Other uses that followed included sporting goods and medical equipment (prosthetics). It was not until the late 1960’s that composite materials were considered for concrete reinforcement. Composites are resistant to corrosion and provide the necessary strength and stiffness for reinforcement of concrete.11 FRP rebars have a high strength to weight ratio and desirable fatigue resistance, making this material attractive to construction, however, the durability of composite bars may be affected under specific harsh environments such as the high pH of concrete, and conditions involving temperature variations and exposure to moisture.10 Vulnerability of FRP bars depends primarily on the constituents and the fabrication method of the composite. The three most common fiber reinforcement systems used in FRP’s are glass, carbon and aramid. Each one possesses different chemical and mechanical properties, and their durability gets affected by exposure to different environments. Carbon and ararnid coI‘nposite FRP bars show higher tensile strengths than glass FRP bars, but their cost is considerably hjgher. Glass I I LI has an adequate tensile strength, is readily availabl 6, and its cost is relatively low when compared With other composites. '2 2.2 FIBER TYPES 2.2.1 Mg Glass fibers are the most widely used fibers in polymer composites. These are various types of glass fibers: E-glass is of calcium aluminoborosilicate composition with a maximum alkali content of 2%, it possesses good strength and a moderate elastic modulus, and is a good electrical insulator.l3 C-glass (also called A-glass) is of soda- lime-borosilicate composition and has a better resistance to chemical corrosion than B- glass. S-glass is of magnesium aluminosilicate composition with very high tensile strength and is able to withstand higher temperatures. AR-glass or alkali resistant glass has a higher zirconia (ZrOz) content, which somewhat improves its behavior under high alkali environments, such as concrete.14 Because of its low cost, B-glass fibers are used no st commonly for the manufacture of polymer matrix composites in construction and other applications, accounting for more than 90% of the total glass fiber production. Glass fibers are isotropic, meaning that they have the same elastic modulus and tensi le strength along the fiber axis and perpendicular to it, due to their three-dimensional network structure. The tensile strength of glass fibers is quite high (3500 MPa [508 ksi]), hilt the elastic modulus is not very high (around 70 GPa [10.2 Msi]) when compared to Carbon or aramid fibers. Consequently, the strength-to-weight ratio is high but the . . . 15 . . modulus-to-wetght ratio 18 only mOderate. This is the reason why the aerosp ace in dus uses other fiber types (the so called advanced fibers). instead of glass Glass fib . er resins within the construction, marine, transportation and other industries Ordinary st 1 ' ee reinforcement has tensile strength and elastic modulus of around 520 We: (75 km) and 200 MPa (29 Msi) respectively. Table 2.1. Comparison of Mechanical and Physical Properties of Different Types of Glass Fibers.'3' ‘4 Bulk Tensile Tensile Failure Type Density Strength Modulus Strain g/em3 (lb/a3) GPa (ksi) GPa (Msi) (%) 2.62 3.4 81.3 E'glass (163.5) (493.1) (11.8) 4'9 2.50 4.6 88.9 5'3”” (156.1) (667.2) (12.9) 5'7 2.56 3.3 09335 (159.8) (478.6) 4'8 70.0 2.78 2.5 AR-glass (173.5) (362.6) (10.2) 3.6 Glass fiber fabrication starts out by melting the raw materials in a furnace heated to 1 540 °C (2642 °F) (Figure 2.1), feeding into electrically heated bushings, which contain a number of holes at their base.15 The molten glass flows by gravity through the holes forming fine continuous filaments which are sprayed with water to cool off. These filaments are gathered together to form a single strand, and a “sizing” is applied bef winding on a drum. Sizing of fibers is provided to protect them from being damagedOT-e handling and to minimize the introduction of surface defects. For I’Cinforceme: purposes, a size based on polyVinYI acetate and containing a resin-coupling agent is used The resin-coupling agent is compatible with polyester, epoxy and phenolic matrix resins The coupling agent is used to bond the fibers and the resin matrix together. Furnace 1” i :1 _ 1 540°C \ RCfiflCr Forehearth :3 Meltcn glass»; ‘ 7 W: _ . e. W»; ‘- . -r‘ ‘v ,“jw :,L" “‘_‘ g - ._. - .. Figure 2.1. Glass Fiber Melting Process.l3 Fibers can suffer degradation when exposed to different environments; this is caused by the interaction between the fibers and deleterious substances in the surrounding environment. Damage to glass fibers by fluids is initiated by physical or chemical reactions between the two. The extent of damage depends on fluid type, fluid concentration, and the composition of the fibers under attack.2 Glass fibers are known to degade in the presence of water, acidic and alkaline solutions.16 Exposure to highly 10 alkaline environments has been found to have the most severe effect on glass fibers Th ' is a “1350‘ concern When utilizing glass fiber reinforced P01ymers as reinforcement f 18 concrete, since the pore water solution that is Present in the micro-voids of concrete has or typical pH of around 135- The high alkalinity of concrete is a consequence of the higha sodium and potassium solubility in the cement matrix. 17 The hydroxide ions (OH') in an alkaline environment attack the primary component of glass (silica or $102) and cause a breakdown in the Si-O-Si bonds in the glass network (Figure 2.2), resulting in fiber degradation and loss of strength and stiffiless.3 I I e I —Si—O—Si— + OH' —Si—OH +SiO' I I I Figure 2.2. Breakdown of Si-O-Si Bonds in Glass Fibers. The network breakdown leads to surface damage on glass and the reaction products may either dissolve or accumulate on the surface of the glass'.14 E- glass/vinylester rod samples were subjected to a 3% ammonia solution at 80°C (176 °F) for 28 days. The results of Raman spectroscopy studies combined with pH data indicated that elevated pH caused a rapid attack on glass fibers.4 The same study determined that ele‘lated pH also causes rapid fiber dissolution followed by interface debonding between the fiber and the matrix. Another study used both E-glass/vinylester and E-glass/polyester Pultruded rods embedded in concrete to study the effects of the high alkali pore solution 0f concrete on the bars. Afier aging in water at 80°C (176 °F) for 14 days it was shown 11 that the reinforcing glass fibers near the surface Of the 1' 0d, Where affected by the alkaline concrete solution, and Were proven to be more brittle and susceptible to damage b mechanical stresses than those: further away from the surfaces Significant Wei ght 108: associated with fiber dissolution Was observed on a study performed on sever-a] fiber/matrix systems utilizing E-glass as the reinforcing fibers after exposure to basic media at different temperatures.’8 Weight loss of the samples increased as the basicin and temperature of the solution increased. In another study, samples made with E-glass fibers and different matrix systems were exposed to a simulated concrete pore solution (pH=13.5) at 60 °C (140 °F) for 9 weeks. The results showed a decrease in mass due to dissolution of the fibers and SEM pictures showed disintegration of the interfacial bonding and erosion of fibers.6 Another study investigated the durability of vinylester and polyester matrix glass reinforced rods and plates by aging in ammonia solution for up to 224 days at 23 °c (73.4 °F), 50 °c (122 °F), and 30 °C (176 °F).19 Gravirnetric and thermogravimetric analyses showed significant weight losses of all samples associated with fiber dissolution when exposed to the highly basic ammonia. solution, with more noticeable effects on the samples exposed to higher temperatures. In contrast, samples exposed to the acetic acid solution for the same period of time showed no weight loss indicating that possible fiber dissolution was not significant. Most of the research related to the durability of glass fibers has focused on the durability of the composite (fiber/matrix) as a whole. One of those studies utilized 12.7 mm diameter (#4 bars) and 19.05 mm diameter (#6 bars) glass fiber reinforced polymer bars exposed to different types of environments and tested in tension to determine 12 strength losses. The bars showed strength losses of up to 64.3% when eXposed to high 0 . alkalinity for 203 days, up to 49. .1 /o for a combination of high alkalinity and freeze-thaw cycles, “P to 765% for a combination of high alkalinity and application of stress, but no significant strength losses were fOttfld for the samples subjected to salt attack. ' A recent study investigated the durability of alkali resistant glass fiber reinforced plastic bars by simulating different exposure conditions at 25 °C (77 °F) and 60 °C (140 °F) for six months and tested for tension.3 The results showed reductions under 8% in tensile strength for the samples immersed in water at 25 °C (77 °F), losses of up to 20.8% and 28.0% for bars exposed to a high alkali solution (pH=12) at 25 °C (77 °F) and 60 °C (140 °F) respectively, no noticeable strength losses for specimens exposed to an acidic solution (pH=3) at 25 °C (77 °F) and for samples exposed to seawater and deicing salt solutions, and strength losses under 6% for samples subjected to ultraviolet rays. A different study dealt with the durability of GFRP pultruded rods by Subjecting samples to two months of immersion in 23 °C (73 °F) distilled water, and to embedment of additional bars in concrete with wetting-drying cycles in water.7 The test results showed that absorption of water produced small reduction in tensile strength and elastic modulus of the order 1 to 7% and 1 to 10% respectively. The bars embedded in concrete produced losses in tensile strength between 6 to 21%, and in elastic modulus between 3 to 11%. 2 .2 - 2 Other Fibers Other types of fibers such as carbon and aramid fibers area also used in fiber reinforced polymer (FRP) bars in concrete. Carbon is a very light material and offers the highest modulus and strength of all reinforcing fibers (230 — 320 GPa [33.4 - 46.4 Msi] 13 6 arbor; fibers 31' . . - es of O and 4.5 — 5.5 GPa [653 - 798 RSI], respectively)” TW° mam VP (coal tar . . . . 'tC PAN carbon fibers (made pnmanly of polyacrylomtnle) and P1 of applications in artetY S shut“ such as car go the aerospace and Sporting goods industrieS. Some parts in the . . . ofl {i‘oet remiotoed epry pitch is its primary component) .14 Carbon fibers have found a v bay doors and booster rocket castings are made of carb composites. Modern commercial aircrafts also use carbon fib 6‘ reiniomed ”meshes. Various other machinery items are made 118ng carbon fiber reinforcement and also in tlle field of medicine the applications include equipment as well as iInplant materials. carbon fibers do not absorb moisture and are resistant to many Chemcal SOlutions, making them . 21 particularly suited to envrronmental exposures. Aramid fibers consist of planar sheets 0f molecules linked together bonding. The proverties 0f the fiber’ and in Particmar the "‘0 qus of elasfib.y h:d’°g°: on the degree of alignment achieved during PrOduction, and therefore ar .cuy, 6pm amid fibers can be of different qualities. DuPont Company has done extensive research on th. fiber and has been able to PTOduce high modulus aromatic fibers commerciall 18 type Of name Kevlar. Kevlar come8 in “"0“ f°rms i“Cluding Kevlar 29. Kevlar 49, 2:11;“ ihe 149 differing in morph°1°gy between cad} “her due t° Pmcessing variables. Kevla:v4: has a higher tensile strength and m°d“1“8(4-0 GPa [580-2 ksil and 131 GPa [19 Msi], 1‘ GSPectively) than Kevlar 29 (3.2 GPa [464-1 ksi] and 83 GPa [12 Msi]. respectively), and thus, it is more widely used for load bearing applications. Kevlar 29 has a higher strain to failure than Kevlar 49 (around 4.0% vs. 2.5% for Kevlar49). Kevlar 149 has an ultra high modulus (186 GPa [27 Msi]).13 KCVIar fibers are used in a variety of products including 14 - s ropes, cables, fabflc g Wi th different . 31011 such as the ones used for bulletproof vests, as reinforcerneflt 1” FRP reinforcement for rubber products such as radial tires for vehicles: '68. - dusm ' ‘ o . 1“ types of resms for use in aerospace, marine, autOmOtive, and Spot‘s 2.3 MATRD( PHASE 2.3.1 Qiem The matrix binds the fibers and particles together in a composite, transfers the load to them, and pI'Otects them against environmental attack and damage due to handling. There are several tYPGS 0f materials used as matrix in a composite such as ceramic, metal, and polymers, the latter being the material used for the manufacture of composites for the construction induStTY- POIYIners have a 10 c . . . . 1 08t’ easy process1blllty, relatively good chemical resrstance, and a 0w Specific graVi . ty, 1"“ they have a low strength, low elastic modulus, low operatlllg temperatures , an . . d a 10W resrstance to prolonged exposure to ultravmlet rays and some Solvents. Became of th e Predominantly covalent bonding, polymers are generally pOor condUCtOTS of heat and elec hicity, Structurally polymers are giant chain like molecules with c , OValentIy bond ed Garb on atoms fonning the backbone of the chain. Two major classes of polymers b can e identified in the production of polymeric matrix COmPosite materials: thermoset and thermoplastic. 15 2.3.2 Thermoset Matrices .33] reactions of 10W 61111 . Omar. When 8.11 - ' I w 01‘g molecular weight monomers, or by homopolimefizanon of kIXOWn Thermoset polymer matrices are formed from the Oh epoxide or epoxy monomer is mixed with an arnjne or C ‘ , , m as powmenzar hardener) along with the application of heat a Chemical process “‘0 1011 occurs.” Polymerization is a four-step process (Figure 23) that Sims out thh the melting of the monomer and forming of a low molecular weight prepolymer (A‘Stage). Afterwards the prepolymer grows and branches (B-stage), the gel POint is reached (Patti a1 cross linking), and lastly the polymer eventually becomes a fillly cured crosslinked glass (C-stage). Upon completion of the 01111118 process all irreversible m'dimension a1 DCtWoz-k structur ' formed. Crosslinking makes sliding of molecules e IS Past one aDOther difficult thus making the pOIymer strong and rigid- The deformation 1) . ehélwor f - 0 thermosets lS controlled by the network structure. The thermal stability and elast' _ . . . . . 1c mOdulus increases with crossi'mk densuy. The Initial Viscosnzy is low . for ease ofpmcessl'ng and their shelf life is limited due to monomer chemical reactwity 16 GEL Pony“. \ A . 13 F'gure 2 3 Polymerization of Thermoset R6511), 1 . . One of the most important properties of tliel‘rnoset polYIners is the glass transition temperature 09.13 The glass transition temperatm'e is defined as the paint at which a polymer tr ansfol’ms from a glassy solid to a mbbery material. When a Polymer liquid is cooled, it contractS. The contraction occurs because Of a decrease in the thelunal vibration of molecules and a reduction in the free VOIUme, meaning that the rnolecules occupy space less loosely. For amorphous polymers, this Contraction continues below the melting Point of the crystalline polymer Tm ‘0 the glass “a“Si‘ion imperative Tg Where the supercooled liquid polymer becomes eXtremely rigid owing to extremely high ViSCOsity, 15 BeYOnd the glass transition temperature the elastic modulus of a POIymer is significantly reduced due to changes in its molecular structure. The value of Tg depends on the type of resin but is usually in the range of 65 to 120 °C (149 to 248 °F)- 17 112:5, oxy, pheno Some of the most common polymer matrices include pOIYBSter’ 6p ' forced 5 fiber rem . . . . as polyimides, and bismalelmldes (BMI).l3 A majority of comI'non g1 and 8% 011 . . between 4 composites have polyester as tile matnx, but p01yesters Shrink . . deu {0‘ uses in concrete . curing and are not very resistant to alkalis and are typically avo Vinylester resins are resistant to a wrde range of acids as we chlorine making them ideal for marine environments, Polyimides have a relatively high service temperature range (250 ' 300 0C [482 - 5‘72 °F1)’ but like other themlosettillg resins, they are brittle.” They are mainly used as high temperature p Olymer matrices in the aerospace industry, and electronics. A major PTOblem With Polyimides is the elimination of water of condensation and solvents during processing. BMI Polymers can have a service temperature between 180 and 200 °C (356 and 3 92 °F), The h . Y ave a good resistance to hygrolhermal effects, but being therm0sets they . e brittle and must be cured at higher temperatures than conventional} epoxies, p OXy 1' eSins are more expensive than 13°13'68th but t1133’ have a better moisture resi Stan Ce 1 ’ OWel- S - a hunk ge on curing, a higher maximum use temperature, and good adhesion , With glass fibers - A v large fraction of high performance polymer man-1x Composites has themo Cry Setting - . - epoxies as matrices. Some problems “nth epoxy resms are that they can degrade in a, e Presence of moisture and elevated temperatures. Moisture plasticizes the polymer 1 d ’ ea ing to swel ling, lower Strength, lower modulus, and lower glass transition temperature 2.3 , 3 Thermoplastic Matrices The other major class of p01YIner is the thermoplastic Phlyrner They are called thermoplastic because they SOfiBr1 0r melt upon heating, and can be re-melted and 18 at weigflt’ hjgl—l me n 50 reformed. They are characterized for having a high molecul solvents- i semi-crystalline or amorphous arrangement, and are Soluble y also can be . e . . . ems, Thermoplastic resins are easwr to fabncate than thermosemng ‘6 cl sh \XBA ‘0 form m ape recycled and posses unlimited shelf life. Heat and pressure are 219'? ‘ them but most often suffer from fabrication Strains from goofing gaolents and other processing-induced stresses.13 Some of its advantages are that they can aChieve a higll toughness, they can be repaired, and a minimum amount of scrap is lefi out during fabrication. Some commonly used thmmOPIaStiCS are polycarbonate’ PEEK, Saturated polyesters, polyamides, nylons, and p01ypr0py1enes - An important problem with 1301)“er matrices iS aSSOCiated With the env' ironmental effects. Polymers can degrade at moderately hlgh temperatures and through moisture absorption. Absorption of moisture from the environment causeS . . 15 Swelling in the polymer as well as a reduction in its Tg- In the PTCSCQCG of fibers bond ed . to the matrix, these hygrothermal effects can lead to severe internal stresses in the co m . . . p 081“: The presence of thermal stresses resultlng from the thermal mlsnlatch between matrix and fiber is a general problem in all kinds 0f comPOSite materials- In pOlyIner matrix compo . SlteS, it is a bigger problem because polymers have high coefficients of thermal expansion 2.3 .4 Fillers Another important component in the composite matrix additional to the polymer is me filler material. Fillers can be added to thermosetting 01' theITIIQplastic polymers to reduce resin cost, control shrinkage, improve mechanical properties, and impart a degree 19 a 'rnprove 10 ' . t’veIY ‘0 1 of fire retardancy. In structural applications, fillers are uSed $6130 1 equitanents 0f the arbonate, and transfer and to reduce cracking in unreinforced areas.22 Clay, . m6 ‘ - t1 ed (1 d on glass-milled fibers are frequen y us epen lng up A as no ally “gated application. Filler materials are available in a variety of form5 «gm gamut. with organo-functional silanes to improve performance and flange 1011. Although minor in terms of the comPOSition of the matrix POIYIner’ a range 0f 1mpm't-Eu'n; additives, including UV inhibitors, initiators (catal ysts), Wetti “3 agents, Pigments and mold release materials are frequently used. 2.4 COMPOSITE FABRICATION Various techniques are used for making glass fiber polymer matn' . X composnes. . . f all in which glasS fiber 6 laid Hand lay-Up 18 the Simplest 0 S at Onto a mOId by hand and windin ' , . g 18 “Other Very versatile technique in which continuous tow or rovmg is paSSed through a resin impr egrlation bath the resin is $1)!an on 01‘ brushed on. Film'l'lent and wound over a rotating or stationary mandrel (Figure 2-‘0-23’24 A rovi . 118 consists of tho f' d' ' 1 filaments. The Wll’ldlng 0f roving can be I . usands o in 1vrdua P0 at or helical, In the founer the fiber tows do not cross over, While in the latter they do. The fibers are laid o I] the mandrel in a helical fashion in both polar and helical windings, the helix angle depending on the shape of the object to be made. Successive layerS are laid on at a constant or varying angle until the desired thickness is attained. Curing 0f thC them-losetting resin is done at an elevated temperature and the mandrel is removed. Very large cylindrical and spherical vessels are built by filament winding- Glass, carbon, and 20 . ester 1. . . y used with epoxy, polyester, and Vinyl esms for aramid fibers are routinel my mac I . ' mm me ' es. Filament Wll'ldln rocess is a 00 used producmg filament wound shaP g p i new to create . 1 'n manufacture of FRP rein forcing bars or construetion app 1 f 'I'l'egularities on the surface of the bar that improve bond perform 1 Tension adjust Resi'mnc<>3l°(l fibers \ . O. Q \\\\-‘ O :.~:.~'_.~ Figure 2A. Filament Winding Manufa cl . n g PrOQ $38.13 Other manufacturing processes include bag molding (Us parts”), stacking and subsequent “mug 0f (”‘6“th prepregs (thin for mall S . ' - - . - eel- [0-04 inch]) of partlally cured resm containing lvlflldll'ectionauy al' 8 (1% 1 . 15 - ed fit) “a” molding, extrusion, thermoformmg 9 and reinforced reaction injectj gee. 011 m0 ['0‘]. . . . . 26 1 6‘01 (Widely used in the automotive mdusth) - 8 (RR ‘ 1M pultrusion is a method that allows the fabrication of continuous fib comesites in the form of sections such as I or T beams and hollow Seet- (F er Ions igure 2-5 35 ° such as 1 hr, or carbon in desired orient - . ) Fibers g ass, Kev anons are impregnated with thermosetting resin and pulled through a heated mold or die. The mold is heated to 2i , d in the curing of the resin- Additional bands of fibers can be in WraPped “finds in the production of FRP rebar to create a bond1ng “thee. 3\ E1roan d s . . \lar “abal- for use in construction purposes- Fiber placement te§\“ x \Q \llat 0fSt1 Q . catalYSt level dye temperature, and pulling Speed are “\BMQ used resins in this rocess are Vi mm“ 0ommonly P ny188ter and h“ h “hm thal are produced ‘15ng ans process include reinforcin “\yfi er QYS‘ The re“ atin s, automo '1 51121118, gro elemen‘s’ gr g b’ 6 “nd anchors dpl‘eSb-e “Some m . 61% window frame sections. 1‘1 663% log rods; “In s , Sheet p171” . g 0 - Heated die . . g ‘ . o o ‘ : -—~ Fiber roving .t “ hue" Resin 58th Figure 25' Pulmsion ManUfacturing Process. ‘3 osrrrag 2.5 MECHANICAL AND THERMAL PROPERTIESQ 0? COM? 25.1 W11 and Elastic MW - Tensi - - . ore failufe' FRP bars do not exhibit yielding or plastic 1 chm/10f bef h ' t unt ' ' _ in Pa behavior of compos1te bars 1s characterized by a lines 1 e135 ti 6 stress stra ultrudee failure (Figure 2. 6) Typical tensile Strength and ClaSti c m du 11.1 S values for {P 22 g1 59 9, carbon and armnid fiber reinforced polymer composite bars ‘ome fractions are shown in Table 2.2. v0 500 a, 400 St :1. Be 2 GFRP ' a; 300 (D 2 ‘3 200 7‘2; C I33. 100 .0. 5 § 0305 0.01 0.015 0.02 0.025 0 03 ' 0.03 5 0-04 0.045 0.5 ‘ - ss-S ‘ . l:igllre 2.6- Typlcal T603113 Stre tram Relatmnship of Pultruded GF ° RP B 60 Reinforcmg Steel. ars ar 23 ' FRP Composite Bars and ’le Properties of Table 2.2. Ten51 Gradfi 6 Reinforcement.‘ 1 0 T ensile SCI-E35; Tert‘sile Elastic MGM“ ‘ MPa 0‘80 GPa (MSi) \l\\\ 483-1600 (70-23 0) 600-3 690 (87-535) 1720-2540 (250—368 ) 517-689 (cfgizleo) \ (75-100) (29) 60.120 . . F b f“ "(31153“? Characteristics of RP ars are dependent on sev al factors such as t f - 0111me fraction, bar diameter, ra e o cur-mg, v twill 27 The 31” ac 3 process- . volume will det i e e s sent per unit 0f fibers pre em] I ctio osite Tensile strength redu ns 0f up to 40% have been found er GFRP comp . an increase in diameter from 9.5 to 22.2 mm (0.375 to 0-875 lhchfiahr When due to ' ed ' ‘mdarchz ’ 611' f FRP bars are, Unhke steel bars, ntfi ‘5 strength prop 168 o (1 Should er. obtained from the n anufactur . inar lay 2.5.2 Shem sue“ due to InterlaI“ R P 318 . - £113 61' I known to be relatively weak in Sh <3 01 F b are - 1y weak P 3 Thus, the {3121th6 ' between the layers of fibers. unreinforced r651“ go‘s/ems interlaminar Shear strength. 24 ' - ‘ direction across the \‘ayerg Orientation of fibers in an off mus Qt . - increase depends upon the de ee the Shear resl stance, me Q can be produced by braiding or filament windi -entation ofl . fiberS- Off-axis fibers can also be Placed me maln m ‘h “3 tram arse to (1 ’ng a continuous strand mat 1n the roving/mat Q Q intro 11:31 Dbl ' 1‘6 Shear b - §°L S t established to characterize 1: chat/101- of [e tend not ye a; articular FRP bars are needed 9 they Should be Obtai ofap ’63! Ineth a" lf‘tbe ods are S bear propelrlbs 50 112 bar Manufacturers. , I 2.5.3 green - - ' Ted to a co 1 s sublec nstant oad - . FRP reinforcing bar over tune can sud deftly fail afier time. This h . e (of ' dufance P Gnomenon tut a time period called the en 13 known as creep (a? cad e is not general] . . - einfot static fatigue) Creep ruptuf y an issue w‘th Steel bars in t . high tern era . S concrete, except 1n exuemely p hues web as those encountered 1n 8 fire. A . - 6 stress to short- f . the ratio of sustained tensfl term strength 0 the FRP bars increaSCS, 6 CI'C time can alg . end‘urance time decreases. Th ep mpture endurat 103 - Q IITeVeI-Sibly ' - . S SuCh as 111% t d fiei iently ath‘irse enVlI‘Onmental cond1t10“ ’ emperature, decrease un 61‘ su c a” wee» or fi§€21fla~mawing . - ° osure, high alkalinity, wet and , a 0 ultrawolet radlatlol'l exp , bl if f endmance times beaten 1 Q . - - tion is currently availa e 0 limited informa cycles. Very hours.1 1 load levels I fit t differe C ep rupture tests conducted on smooth GF R P b ars a re arithm the 103 ' d“ t that a linear relationship ex1sts between creep mpmfe streflgt in ma e 25 - O h , The ratio of stress level at Q ftirne for times up to nearly 1 0 ours ree1) ru 0 131; , . 111 than 50 ears) were linear“ Elma W9 after 500,000 hours (more Y ““lated to b e , tigation of creep rupture has been reported. flu Q1? Another “Wes W Kent f initial tensil e strength retained follo “Rh: centage o W eh ~ The per 0 a “\QBK \ gafithrn of time, reachlng 3 W11 116 0f 55 /° at an eXtrapolit 3 Wm“ ° (5 . We 2 5 4 W The coefficients of therflla1 exl’ElIISion of FRP b 313. directions depending of) the types 0f fiber Val}, in longiwdvh; transverse d 1’681'12, and 1 . - . . t of th 1 . V0 1.11116 fiactl¢ fibers The longltudmal coef‘fi01611 Erma eXPaHSIQ . 5 . IS dominated by the pI‘Opel of fibers while file transverse 006 Clent Is domlnated by the resin 31 Table 2.3 . Coet‘fic'u‘:fit of Thermal EXPaDSiOH 0f FRP Composite Bars, SW Concrew“ W GFRP [\Z‘Efi Lo W W -2 to 0 -6 to -2' “'7 \ lnogl'6 /°C (3. 3 t0 5.6) (‘4 t0 0) (-3.3 to -1 A3 (4 to ( 1 0 *5 M?) W 2.]to7-3 60 to 8Q (11-7 to 12.8) (41 to 58) (33.3 to 44.43 Transverse 1 0 '6 /°C ( 1 o *3 /°I:‘) (4 to 26 \- 5 5 Bond Strengm 2. - Bond performance of an FRP bar is dependent on the QQSX mm o o . NIH . process and Incohanlcal pmpeflles 0f the bat use“: 3116 'h\so Q faetunng Q . . . 32-36 When anchoring a reinforcing bar i ' n c h oofidltions 0 q“ 311quer by the adheSiOn reSiStanCe of interface (Ch ‘ ‘ n ‘1 e ‘3“ ' - Q'n me can b of the interface agalnSt $11p, and/or mechanical interl I Mud), . e interface. In FRP bars, it is postulated that the bend fore . _ . Is to the reinforcm g fibers, and a bC’11d Shear failure in 01% ”any ' . res ‘ , bonded deformed bar is subjected to 1nCreaslng tension, u) I” IS also c-é'(17}1¢<,1<,~1,-0r1 be 011 possible. ”‘02 can a the surrounding concrete breakS dOWn, and deformati “”6611 the bar and S 0 e b n the surface of bar cause inclined Contact forces betWeen t ar and surrounding 00 11¢er 6. Unlike reinforcing ppears not to b - . steel the bond 0f FRP rebars a e Slgmficantl ' rete , y lnfllien the “one ced by - ' ed th cornpressWe Strength, PTOVId at adeqUate concrete co Ver 37-39 exists longitudinal Splitting. Environments that degrade the POIymer resin or fi‘oetlresin inter-faQe ar e likely to - th degrade the mud Strength 0f composne bars. Direct pu\\0“‘ “18‘s are e jthost Common ““1 estigat§ (1 th methods for testing bond of F RP bars to concrete. One sway e effect 0f ' ' u natural environmental exposure c‘m‘htlons on Glass FR? 9 40‘ M . 2 Years“ onal F RP At elevated no signific ion . . . ient tenlperann-es, m6 incompatiblllty thWeen the transverse waffle Ct]- . Unidlre betWeen composite bars and concrete may influence bond Strength . 6 to 8 31151011 bars used in concrete typically have a transverse coeffi (:1.th of thema] exp 27 er than that Of concrete. During temperatures variationS 3 eat ti 1116 gr thermal ' ° the concrete 3 cause s \ittin k 42 ex . .9rnatch stresses Withm m V n g crate s. ~44 Damon 1 . S 0nd between concrete and reinforcemem a - degradation of b y result In teSVOnse' a“ Occurrence of splitting cracks under mermal 1Q - d uch as type of FRP relnforcernent, type of Conctete pr “finds s 3 e, smut a1 (’l’CI [1163 ()1 ”I ’SCCtiOII ‘ , Emml . - . S geometrical PT 6 cross l’le Stud: Q Cf when GFRP pullout s . tea} “in . er could occur Peczmcns \K» ed 013 g actmns mid 00V 1 C I‘ . 45 GF‘ RP bars etc 0 k1,) gradlentso In nether study, were CXPOS 31141th g ofconcrete . . ‘ . . to to twp E environments in concrete (Wlth and wlthOUt arti fielall y We’ elevat -tefnpet- [we dw . anll’e alkahnity), and foun d both negative and POSitiVC impact 0“ pullout Strength over a sho rt Pen'Od of time.32‘ 34, 35, 46 2,6 MECHANXSMS or: GFRP COMPOSITES DEGRADATION 2.6-1 9,1201% Even though glass FRP lS Immune to corrosion attack, it is prone tQ deterioration in the alkali environment of concrete, There are also concerns With the dab” _ aging effects FRP‘ of ultraviolet exposure, elevated temperatures, and moist‘xc 0“ 935‘s y investigators . . li Degradation 2 6 2 Alka 051 interest ‘0 . ' ' m The envn-onmental condition that has attracted titre in outdoor concrete . . . . found concerned With FRP bars 13 the highly alkaline pore water und 13,6, making ' ' 0 structures. Afier concrete has been fully hydrated, its aVerage pH 15 ar 'fy the t0 V611 it a highly basic (alkaline) medium. Extensive research 12 as b 3611 conducted 28 ° ' ass F rods. Since COmPOSite rod 6606': of this enwronment on g1 RP S are ”let. ° - be simuhied ‘ ' material, long term effects have had to Go‘,n_3tru<:t10n dIOdS such as irnrnersion in. high temperature (60 °C “4% m6 (3 ' ed fr - §\\‘ K\\§K\~ \ufions Based on me results Obtalfl om these Investiga “\Q 80 ° ka {\Q . posure to high—a1 1i environments can decrease th “S 61» o 12 ens. (GFRP) rods by as much as 64-3 A) 3 although Partiq “% cording to differences in test metlwds' Results are a 150 ac ’ \2 condition, the resin used to protect and blnd the fibers, Ghendém a0, a rhe high alkalinity of conerete environment is a Co l'ISOQHence of the . ~ Sodiuxn and PotasSi‘lm hydroxides Present in . solubihty hydrated c en t mam" hydroxide ions (OI-1') in an alkallne environment attack the pri mary compone‘»m (silica or Si02) and cause a breakdown In the Si- O-SI single bond fomling molecular structure (F igul’ e 2.7). T1118 results in fiber deterioration and loss Qf t s It /— ‘ NaOH Ca(OH)2' \ r / OH +Si0'3i0‘f ‘ CaLOHn- 46H- / Figure 2,7. Alkali Attack in Pultruded GFRP Bars 29 One study found that E-glass GFRP rods exposed to basic envi 1' Onments suffer fiber dissolution leading to loss of weight and decrease in strength (’me bars '8 Since th - e degree and rate of degradation depends strongly on the amo “In of chemicals that ar e able to reach and attack the longitudinal fiberS, the resin that Protects and binds fibers togfitl'ler plays a very important role in this aSPeCt- Different matrix mat . 1 ma 3 are common] y used to protect fibers, including Vinylester and P0 lyester-baSed . r eSInS- The Federal Highwa y 8 . - . decks. Tests were conducted in an alkaline enVII'OHment on two sets ofgmtings using E- glass fibers, one set with a polyester matrix, another with a vinylester matrix. According to the study, composites with polyester resins are less resistant to alkali attack and Showed very rapid deterioration when compared to composites with Vinylester resm. Three-point bending tests on grating samples after 160 days Of eXposure showed strength . - 0 ~ . 8 reductions of up to 80% for polyester resm, and up to 25 A: for Vlnyl est er resm- Another study supported this observation when the outer surface of . . GFRP Samples contaimng pOIYester resin matrix showed a higher degree of de d ‘ gra atlon co mpared to . V1, . . , , samples w1th nylester resm. This degradatlon process Involved formation of gel . . . \ 17cc . accompamed by swelling, followed by blistering and eventual dissolution matenals 01“ ”I e ~ . some cases.6 res", "1 The rod diameter also influences the rate and degree of degradation of g1 aSs FRP barg exposed to alkaline environment. One study showed that the larger the dialnet er of re ' ' ‘8 (ls, the smaller the percentage welght loss dunng exposure. In the case of a thicker “)6, the average distance of fibers from the surface of the rod increases the fibers thus 30 become more protected by the matrix and the PH rise in the center 0f the rod is expected to be more gradual. 2.6.3 Degradation Due to Other Chemicals Similar to the alkali attack, Salt attack will mainly occur due to the presence of OH' and Cl' free ions in the SOlmion‘ Other ions Present in the solution could to a lesser degree, react with the fiber and matrix. The C1- ions are not as damaging to glass as is the OH" iOIIS, however C1- ions can penetrate the malfiX, causing microcracking and fracturing of the matrix which accelerates moisture diffusion as well as debonding of fibers. Debonding of fibers, in turn, Will I‘CSUIt in the loss of strength of the bar.3 Tests performed on bars placed for 203 days in a capped PVC pipe filled with salt solution consisting of 97% water and 3% sodium Chloride showed no significant loss of tensile strength (some exposed bars even carried a higher load than unconditioned bars).l - for the ma'ori Strength and stiffness losses J ty 0f bars Were Under 1 0% and 8%, - - thin 7.5% and 6°/ 1' ti res ectively, and gains were WI ° espec Vel , . p y These r11mor losses in strength and Stiffness show that glass FRP rods are suitable fol. COnStru . S d d k - 1d . ed to deiCing chemicals (e.g.: pavements 01' b“ ge CC 3 in C0 regions). corrosive environments such as m marine SWCtureS, and those eXpo 2. 6.4 W A primary cause of deterioration of GFRP’S is the diffusion of moisture and Othel- coI‘rosive solutions into the matrix, WhiCh can damage the matfi" as Well as the fib ers. Therefore, moisture absorption and associated changes in material properties must be 31 taken into consideration. Moisture absorption of a composite can be defined in terms of two parameters: maximum moisture content or saturation moisture and mass diffusion coefficient Maximum moisture content is the moisture level in the composite that is reached asymptotically after a long period Of time. it is dependent on material type and temperature and type of the envirornmol‘lt-3 EXcessive absorption 0f water in composites could result in significant 1035 Of strength and Stiffiless. Water absorption produces changes in resin properties and could cause Swelling and Warping in composites.22 There are, however, resins which are formulated to be mOiStUI'C-I‘esistant and may be used when a structure is expected to be wet at all times. In cold regions, the effect of fieeze-thaw cycles must also be considered. Typically, for glass FRP’ losses in tensile and flexural Strengths of 10% or more - 7, 4 . may be expected after a few months of exposure to moisture 7. 48’ although some studies - - 49 . indicate that such losses may be negligible. One particular Study Showed that after 60 - ° - les for 12 16 and 20mm f mmersron and wet dry eye , , (0.47, . . days 0 1 O 63 and 0.79 mch) diameter bars, exposure to moisture had a small effect on the Ultimate t . Young modulus and Poisson ratio. The reduction in Strength Was , - - . ’ and 20 respectively, for the three bar diameters, while the elastic moonlus decreased /o, y 6: 9 an 0. 5% 7 The same study concluded that even after 60 days of immCrsion in d ' w at GFRP bars had not reached a saturated state, and Were continuing to absm-b wat:a :he a Pro gressively increasing rate. Also, the absorbed water produced Small reduction a tensile strength and elastic modulus of the order of l to 7% and 1 to 10%, TCSPECtiVely’ and the stress-strain curves remained linear after 60 days of immersion in water. 32 f9 Tests performed on GFRP bars made with alkali-resistant glass fiber showed that submersion in water resulted in a measured reduction in tensile strength of 7.3% and 5.9% for bars with polyester 311d Vinylester mamoeS, respectively. Limited changes in the elastic modulus were observed, and these chmges were Within 5% of the initial values? 2.6.5 W Many composites have good to excellent Properties at elevated temperatures and do not burn easily. The effect of high temperature is more severe on the resin than on fibers. Resins contain large amounts 0f carbon and hydrogen, which are flammable. - ' ° the develo ment 0f more fire- ' ° 50 Research 18 contlnurng on P resrstant regms, Tests conducted in Germany have shown that E-glass FRP bars could sustain 85% of their room-temperature strength, after half an hour of eXposure to 300 0C (572 °F) temperature ‘ tensile stren .51 ' ‘ when stressed to 50% 0f the“ gth WhIIC this performance is better than that of prestressing steel, the strength loss increases at higher temperatures f . all7d approaches that of steel. The problem of fire 01' a concrete member reinforced With FRP co . - - m osi is different from that of eompOSite materials subjected to direCt fl re. P tes ' te tthe FRP fr ‘ this Case, the Concrete serves as a barrier to Pro C om direct contaCt with flan, th es. HoweVer e mechanical properti . . es the FRP may change sigruficantly. It 18 therefore recommended that the of user as the temperature in the interior of the member increases, Obtain information on the performance of a particular FRP reinforcement and resin Sy Stem at e1 eVated temperatures when potential for fire is hi gh.22 33 [GS re The use of FRP reinforcement is not recommended for structures in which fire resistance is essential to maintain structural integrity. Because FRP reinforcement is embedded in concrete, the reinforcement c‘ehtlot burn due to a lack of oxygen; however, the polymers will sofien due to the excessive heat. Beyond the glass-transition temperature Tg’ the 6135“ modulus Of a P01Yl'ner is Siguficantly reduced due to changes in its molecular structure. In a composite material, the fibers (which exhibit better thermal properties than the resin) can continue to suPport some load in the longitudinal direction. However, the tensile properties of the over-an composite are reduced due to a reduction in force transfer between1 fibers through bond to the resin. Test results have indicated that temperatures of 250 °C (480 °F), much higher than the glass-transition temperature, will reduce the tensile strength 0f GFRP and bars by more than 20%.52 Other properties more directly affected by the Shear transfer through the resin, such as shear and bending strength, are reduced Significantly at temperatures above the Tg.53 For FRP reinforced concrete, the properties of the polymer at the S If bar are essential in maintaining bond between FRP and concrete, At a tern u ace Of the to its glass transition temperature, however, the mechanical properties of th Peramre Close significantly reduced, and the polymer is not able to transfer Stresses from th:p01ymer are . ' . 0011c]. the fibers. One study earned 011tW1th bars havmg glass transition temperatur etc to es 0 ' ' of 60-124 C (140-255 0F) reports a reduction in pullout (bond) Strength of 20 t 4 0 0% a O O t temperature of approximately 100 C (210 F), and a reduCtion of 80 to 900 a /° at a telnperature Of 200 0C (390 °F)'S3 Another study involving FRP reinforced b rep 01.th reinforcement tensile failures when the reinforcement reached temperatu res of 34 250-350 °C (482-662 °F).54 2.6.6 Wlnavioletw The ultraviolet rays present in sunlight can damage composites. These rays cause chemical reactions in a polymer matrix, Which can lead to degradation of its preperties. Although the Women can be solved With ihe introduction of appropriate additives to the resin, this type 0f damage is “Qt Of concem When FRP elements are used as internal reinforcement for concrete structuI es, and therefore not subjected to direct sunlight.22 Extended exposure of FRP bars to ultraviolet rays and moisture prior to their placement in concrete could adversely affect their tensile strength due to degradation of the polymer constituents and resins. Proper ConStI'UCiion PraCtiCeS and resin additives can ameliorate this type of weathering problem Significantly. Resins, in general, will be affected by UV unless adeqUate pr o te t° . d d c ion 18 prov1 e by additives or coatings. In turn, the composite properties Would 313 b 0 6 affected , most in compression, shear, and transverse tension.8 y 35 3.1 CHAPTER 3 EXPERIMENTAL PROGRAM SAMPLE PREPARATION AWUIONING Tensile tests and pullout (bond) tests Were performed on several pultruded GFRP bars proVideCi by four different manufacturers to determine the effects of different properties (fiber volume fraction, matl'lX 00111190310011, bar diameter, coating treatment) on degradation due to alkali exposure. These bars Were made out of continuous E-glass fibers with either polyester or Vinylester matrices. The experimental plan for tension and bond tests are summarized in Tables 3.1 and 3.2, respectively. Table 31 . Tensile Strength Test Frog-an] 36 Matrix Fiber Volume Conditiom'n f Manufacturer Bar Fraction b g 0 ' y Uncoated B Diameter Weight, % ars NT 6.3 5 mm Wter 70 Control; 14, 28 and (0.25 in); 42 day Immorsion 9.3 5mm and Wet-dry cyeles. (0.375 in.) E’fi" 9.35% Vinyles‘e’ 60 Control; 14, 28, and (0.375 in.) 42 day Immersion and Wet-dry Cycles. fir 8.0mm Polyester 60 Contrm (0.315 in.) day Immersion and Wet-dry Cycles 1V 9.35mm Polyester 50 Comm (0.375 in.) Immersion and Wet- dry cycles \\ Conditioning of Coated Bars “Scam" 28 day ersiOn and Wet- ; 42 day Table 3.2. Bond Test Program Manufacturer Matrix Coating \ Curing Unfited Coated Control Aged Control We Epoxy w. 21 days in 21 days. of I l 5% Silica 20°C (68°F) immersmn Fume lime in60°C saturated (140W) water water 11 Vinylester None 2 1 days in 21 days of 20°C (68°F) immersion lime in 60°C saturated (140°F) Water water B s ranging from 6 35mm to 9.53mm (0.25 in. to 0.375 in.) diameter provided ar . b four different manufacturers were considered in uncoated condition; in addition, some 3’ bars received epoxy coating- Bars from man‘JfaCtul‘ers I and II had deformed surfaces and were produced for concrete reinforcement Bars from manufacturers III and IV had smooth surfaces and were pultl'nded for other applications, Bars were tested in tenSion either in control (CTR) Cofldition prio r o . after two weeks of storage in 50% relative humidity at 20°C (68°F), 0r afieraZ'f:gIn& durations of continuous immerSion (CS) 0r wet-dry cycles (WD) in an alkaline 80111?th The alkaline solution consisted of 16.6 g/L of Potassium Hydroxide (KOH), 2.36 g/I:0(: Sodium Hydroxide (NaOH), and 2.5 g/L of Calcium Hydroxide (Ca(OH)2) heated to 60 0C (140 °F). This solution was selected to simulate the pore solution of concl'ete and Provided a measured pH value of 13.6 i 0.1. Figure 3.1 shows the continuous immersion 37 of bars in the solution. In accelerated aging (accelerated test method to predict long—term performancels’ '9), each wetting-drying cycle involved 16 hours of immersion in a 60°C (140 °F) alkaline solution followed by 8 hours of air-drying, Different samples were subjected to 14, 28, and 42 days of continuous immersion or repeated wetting-drying cycles in 60°C alkaline solution. Close to six replicated tension tests were performed for each distinct condition. Figure 3.1. Aging Tank with Alkali Solution Bond tests were executed in order to determine the effects of alkalinity in concrete in bond strength characteristics between GFRP bars and concrete, These tests were performed either in unaged condition after 21 days of moist curing in 20°C (68°F) lime saturated water followed by 7 days of conditioning in 50% relative humidity at 20°C (68°F), or in aged condition after 21 days of moist curing followed by 21 days of immorsion in 60°C (140°F) water and then 7 days of conditioning in 50% relative humidity at 20°C (68°F), Four replicated unaged tests and three replicated aged tests were performed for each distinct condition. 38 3.2 TESTING PROCEDURQ 3.2.1 Tension Tests Tension tests similar to testing procedures previously employed by several researchers“ 6' 22 were performed on 864mm (34 inch) long bars. Steel pipes 19 mm (% inch) diameter and 178 mm (7 inch) long With a thickness of 3 m (1/8 inch) were placed at the ends of each bar filled With epoxy (“3-932 epoxy resin from Adtech Corp), and left to cure for 10 days at 22 0C (72 °F) and 50% relative humidity Prior to testing (Figure 3.2). A servovalve‘comroned hydraulic test system (Figure 3.3) With circular groove jaws was used for tension tests which were performed at controllfid deflection rate of 0.01 mm/sec (0.00039 in/sec). Loads and deflections were measurfid throughout the tension tests. \ GFRP Bar (gauge length = 508.0 mm) W eluezJeL, i Steel pipe filled with epoxy (diameter = 19.0 mm) l *— Figure 3.2. GF RP Bar Tension Test Specimen 39 Figure 3.3. Tension Test Set-Up 3.2.2 Pullout (Bond 2 Tests The bond test set-up is presented in Figure 3.4. 152 mm diameter by 152 mm height (6 inch diameter by 6 inch height) cylindrical plastic molds were used to embed 610 mm (24 in.) long bars in concrete, with an embedment length of 51mm (2 inches) at the center, and debonded 51 mm (2 in.). The bar was located at the center of the concrete cylinder; the bar was debonded fiom the concrete over a length of 51 mm (2 incheS) at both ends of the cylinder (Figure 3.5). A metallic tape covered with grease was used to break the bond between the bar and concrete at two ends (Figure 3.6). 40 FRP Bar Concrete Cylinder m) Bonded lengh = . ‘ (diameter =152-0 m Figure 3-4. Diagram of Bond Test Sample Figure 3.5. Setup of Bond Test Specimens Prior to Pouring of C011 Crete 41 of Bar Figure 3.6. Detail of Development Lengtt‘ and Debonded Ends Normal strength concrete With an average compressive strength 0f 11.9 MPa (3500 psi), a slump of 76 mm (3 iHCheS) and 3% air content was used i0 935” he bond test specimens. An electric concrete vibrator with a rounded head Was used to eons ornate “‘6 concret ' 'd 1d ' ' wiih e 1nsr e mo 3. Bond specrmens were cured In 22 °C (72 °F) for 21 days, only 76 mm (3 in.) of the exposed segment of the bars imrners ed . t r A steel tube in wa 6 - filled 'th o si ‘ ' ' wr 313 x‘)’ i mllar to that used in tensron tests) Was attached to the free end of each bar for pu\\out testing Two steel frames were placed a1- ' Olmd the concret - e to provrde the gripping mechanism to the bond samples during testing. Pullout tests W ere Perfonned in a servovalve-controlled hydraulic test system (Figure 3.7) at a control] e d . rate of 0.01 mm/sec (0.00039 in/sec). IspIaoemen t 42 Figure 3.7. Bond Test Setup Ultimate shear bond strength (Tb) was calculated as follows. where, Pmax = maximum load D = bar diameter L = embedment length 43 CHAPTER 4 EXPERIMENTAL RESULTS AND ANALYSES 4- 1 TENSILE STRENGTHQF GFRP BARS - resented in ~ - - ‘ tion are p The raw tension test data generated in this inves’tlga ' ' te that - 1e 4.1 indica Appendix A. The mean tensile strength values presented in Tab in Warm alkaline solution accelerated aging through W et-dry or continuous immersion fier generally causes a drop in tensile strength. The percentage losses Oi tensile Sticng‘h a accelerated aging (Tables 42(3) and 42(1)» are, however, QUite variabie- WNW some.) GFRP bars experience losses as much as 42% in tensile strength, others 1"“ng tam“ their tensile strength afi er aging. Table Alb/16:21.13 Tensile Strength Values (MPa) Prior to and After Aging. (a) Uncoated GFRP Bars ‘Manufacture Unaged 14 days of aging \ 28 days OEngg \ Control Immersio Wet-drmimersio Wet-dry 1 747.9 510.8 542.01 473.4 LT744 11 653.1 578.1 588.9 \ 524.1 111 585.0 3724 ' 749.2 ‘ 381.6 1v 384.4 V 364.9 w ‘4‘ V v 44 Table 4.1. (cont’d). (b) Coated GFRP Bars lfianufactureri Unaged Control lmmersio Wet-dry I 740.0 H 677.5 HI 682.6 Table 4.2. Mean Percentage Dr0p in Tensile Strength A fl Aging er - (a) Uncoated GFRP Bars anufactur \4 da sofa 'n 28 da 3 f '\ ° y g g y 0 aging r42 days of aging W t- 1mm ' -\¥ e dry ersron Wet dry I ersion Wet-dry 1 \ 29 .2 24.9 34.2 33.9 \ 19.7 20.9 ‘iY 34.8 \ 40.2‘ 45 Table 4.2. (cont’d). (b) Coated GFRP Bars IV 12.3 Manilfacturer 14 days ofagifng 28 days of aging 42 days ofagmg Immersion Wt-dry [Immersion Wet—dry WW _ I 1.. X 27.5 \ 20.3 “— H l I” \ “- .. \ 19.0 GFRP bars from maIIUfaCturer I eXhibited the higheSt levels of t ‘ h ensile “egg ° - (enSiiC before aging in both Imcoated and coated conditionS, but Suffered major (055 of strength upon aging (Figures 4.1(a) and 42(3)). Coated and uncoat ed GFRP bars 0f manufacturer 1 \ost about 40% of their tensile strength after 42 days of agi 11g. In the case of GFRP bars from manufacturers II and III also major losses of tensile Strength Occurred after aging (Figures 4.1(b) and (c) and 4.2(b) and (6)). GFRP bars from methanol” ers 2((1))- however, largely retained their unaged tensile strength afier aging (Figures 4 1(a) ' and Iv, 4. 46 Tensile Strength, MPa o§§§§§§§§ -1—--— EH5” ElElE Aging Period, days 0’) Manufacturer II. Figure 4.1. Mean Values and Standard Errors of GFRP B - Results. ar Tensfle Strength Test 47 Aging Period, days (0) Manufacturer 111. Aging Period, days (d) Manufacturer IV. Figure 4.1. (cont’d). 48 o\° d O h D E. U) c 2 (0 £2 '5 a“: ,— «03:58.8 th Drop. % O M U‘ Tensile streng Aging Period, days (1)) Manufacturer II. Figure 4.2. Percentage Drop in Tensile Strength (Mean Values Errors) of GFRP Bars After Aging. and Standard 49 so °\° 45 d 40 9 Q 35 5 30 cc» 25 3 2 w o g 15 g 10 l- 5 o Immersion Wet-dry 'mme‘sion Wet-dry Aging Period, days (C) Manufacturer 111. 50 l o\° d. e 30 0 l E. 20 T E 10* 1 “Uncoafai 25 lewd a) A. -, 0’ lmmersim Wet-dry ’- -10 ~ 28 -zoi Aging Period, days ((1) Manufacturer IV. Figure 4.2. (cont’d). 50 4.2 STATISTICAL ANALYSIS or TENSILB STRENGTH T5571: 55w 75 The Analysis of Variance (ANOVA) technique was used for statistical analysis of results towards assessment of the effects of different variables on resistance of GFRP bars to alkali attack. The first step in analysis of Variance of results involved assessnlsint of the effects of aging duration and type (Wet-dry vs. continuous immersion) and coating (coated vs. uncoated) on percent 1088 of tensile strength due to aging of GFRP bars from of 1" each manufacturer separately. The results are presented in Table 4.3 in the form values representing the probability 0f error in the conclusion that there is a . , . s Slgmficant effect of a particular vanable (e.g., aging duration). P—value . mice. { S'ngtfic o 0‘ GFRP bars The general trends in Table 4.3 indicate that aging duration and coalmg indicate that the variable has a statistically significant effect at 5% “516‘ - of the have statistically significant effects on their alkali resistance because mOSl corresponding p-Values fall below 0.05; the effects of aging type (Wet-dry VS COfltlfluous immersion), however, are not significant. Pairwise comparison of the res . , , IJlts ' aging durations Indicated that, in general, aging beyond 28 days did not at different . . . {in - . additional loss of tensile strength. F1 gures 4.3 through 4.5 present the tr Se Signlficant d3 . of least square means) in effects of different factors on the drop in ten (In the form Si] . e Siren GFRP bars due to aging, 3th Of 51 Table 4.3. P-values Associated with Different Factors Analyzed Effect GF RP Bar Manufacturer 'il Aging Duration W \ II\111\ 0.000 \ 0.032 Type of Aging 0.051 ‘ 0.288 0.270 \03tsj Bar Coating 0-005 41 33 ‘ 25 ‘ Drop in Tensile Strength. % 17 14 28 42 Duration of Acclerated Aging, days (a) GFRP bars from manufacturer I 44.0 00 g» o: 27.5 Drop in Tensile Strength % u u o 22.0 I 14 28 . Duration of Acclerated AQ'nQ. days (C) GFRP bars from manufaCturer III Figure 4.3. Least Square Means and vs, Aging Duration. 0.118 i 0.04:] N N 01 I Drop in Tensile Strength, % so 3 U1 0 l l 3.0 28 ‘7' 4 ' da Durati (Jn of Accbrated Mm, ys (b) GFRP bars from manufacturer H Standard Error of % Drop in Tensile Strength 52 00 O) T l ' 22 3% 34”— *_ 20 E— E 0 0 23 30 — ti; ‘8 :2 .9; 8 23 _. ‘ 8 193 19 16 C __ 'a 26 - ‘2‘: 9 9 14 0 24 .. D 22 1 if‘ immersion Wet-dry _ 12 inmersion wet—:gfion Accelerated Aging Condition Accelerated A95"g CO 1' II manufac (a) GFRP bars from manufacturer I (b) GFRP bars from 42- , #fl 13-0 o\°_ o\°_ 14.6 as. _ E 5.) 5 11.2 a .53 8 2 o o 7.8 2 so i .2: Q Q g g 4.4 24 1 1 — 1.0 Immersion Wet-dry AcCelerated Aging Condition (0) GF RP bars from manufacturer 111 (d) GFRP bars [1‘ Q - a"Ufactur F rgure 4.4, Least Square Means and Standard Error of % Drop in e;- I V vs. Aging Condltlon. ensue S ngth 53 w .‘1 O r r 22.0 °\° 338— ‘ °\°- 19'4 5- 5 g 2’ g 302— ‘ g 168 a) m ' .8 3.? g _ ‘3 f3 26.8— g 14.2 .S E 8 — 8 5 23.4 5 11-6 20.0 ' 4,-— Coated Uncoated 9.0 Bar Classification Coated .uncgfd Bar Class'fic‘ fac (a) GFRP bars from manufacturer I (b) GF RP bars from manu 20.0 o\° 5: 14.5 E; :3 a) .32 9.0 U) C O l- E g 3.5 o _2-0 4 #14; Coated Uncoated Bar Classification (0) GF RP bars from manufacturer IV Figure 4.5. Least square Means and Standard Error of % Strength vs. Bar Coating Condition. bro!) in Tensi 1e Given the results presented above concerning the effects of differem v s a“able S percentage loss of GFRP tensile strength due to aging, the next step in Statistical on . o o atlalySiS involved: (1) combination of Wet-dry and continuous lmmel‘SiOn aging co (1 n ition irrespective of the aging type; and (2) exclusion of 14 day aging results and combinat' 1011 54 of 28-day and 42-day aging results irrespective of the aging duration. The combined data on percentage loss in tensile strength of GFRP bars due to aging were then subjected to multiple-variable analysis of variance with fiber volume fraction, bar diameter, matrix type and coating (i.e., coated vs. uncoated) as the key factors. Analysis of variance of the combined data indicated that, at 5% level of significance, fiber volume fraction is the key factor influencing the percent drop in tensile strength of GFRP bars due to aging. The test data was insufficient to draw conclusions regarding the effects of matrix type (polyester vs. Vinylester) and bar diameter, and the effect of coating was not statistically significant (at 5% level of significance) in this analysis of combined data. The p-values are presented in Table 4.4; Figures 4.6 through 4.7 show the trends (in the form of least square means) in the effects of different variables on percent loss in strength. Table 4.4. P-values for Analysis of Combined Data. Effect Analyzed P-value Fiber Volume Fraction 0.000 Bar Diameter Insufficient data Coating 0.1 l 7 Matrix Type Insufficient data 55 41 l I I 31— __ 21— — Drop in Tensile Strength, % 1 1 l l l 50 60 70 Fiber Volume Fraction, °/o Figure 4.6. Least Square Means of % Drop in Tensile Strength vs. Fiber Volume Fraction for all GFRP Bars Tested Afier Aging. 34 I I o\° 5:32— — 8’ 930— ‘ a 928— "T E £26— - .E 024'— "‘ 2 022.. _ 20 l l Coated Uncoated Bar Classification Figure 4.7. Least Square Means of % Dr0p in Tensile Strength vs. Bar Coating Condition for all GFRP Bars Tested After Aging. 56 4.3 MICROSCOPIC ANALYSIS OF GFRP BARS Scanning Electron Microscope (SEM) images were obtained from cross sections of aged and unaged GFRP bar samples to identify damage of fibers. Images demonstrate particular severe fiber degradation near the surface for all bars after exposure to alkaline environment (Figures 4.8-11, (c) and (d)). Damage to fibers near the surface of the bar is more noticeable on bars with polyester matrices (Manufacturers III and IV) than in bars with Vinylester matrices (Manufacturers I and 11). Images taken from fibers near the center of the bar reveal more extensive damage in aged bars with higher fiber volume fractions (Manufacturers 1, II, and III, Figures 4.8-4.10 (a) and (b)), while fibers from Manufacturer IV (fiber volume fraction = 50%) appear to be unaffected by exposure to alkaline solution (Figure (4.11 (a) and (b)). 57 (a) Near Center of Bar, Unaged Condition (b) Near Center of Bar, Aged Condition (c) Near Surface of Bar, Unaged Condition (d) Near Surface of Bar, Aged Condition Figure 4.8. Scanning Electron Microscope (SEM) Images of Glass Fibers in Bars from Manufacturer I (Vinylester Resin, Vf = 70%) at 350X Magnification. 58 (a) Near Center of Bar, Unaged Condition (b) Near Center of Bar, Aged Condition (c) Near Surface of Bar, Unaged Condition (d) Near Surface of Bar, Aged Condition Figure 4.9. Scanning Electron Microscope (SEM) Images of Glass Fibers in Bars from Manufacturer 11 (Vinylester Resin, Vf = 60%) at 350x Magnification. 59 (a) Near Center of Bar, Unaged Condition (b) Near Center of Bar, Aged Condition (c) Near Surface of Bar, Unaged Condition (d) Near Surface of Bar, Aged Condition Figure 4.10. Scanning Electron Microscope (SEM) Images of Glass Fibers in Bars from Manufacturer III (Polyester Resin, Vf = 60%) at 350X Magnification. 60 (a) Near Center of Bar, Unaged Condition (b) Near Center of Bar, Aged Condition (c) Near Surface of Bar, Unaged Condition (d) Near Surface of Bar, Aged Condition Figure 4.11. Scanning Electron Microscope (SEM) Images of Glass Fibers in Bars from Manufacturer IV (Polyester Resin, Vf = 50%) at 800X Magnification. 61 4.4 BOND STRENGTH OF GFRP BARS The raw bond strength test data generated in this investigation are presented in Appendix B. The mean bond strength values (Figure 4.12) indicate that bond strengths of deformed bars from both manufacturers (I and II) immersed for 6 weeks in 70 °C water (aged) were comparable to those of unaged bars (immersed in 20 °C water). Coating also did not strongly impact bond strength. The effects of aging and coating, however, become apparent if one looks at the mechanism of bond failure. 10.00 - 30.00 25.00 - 1 1% . ‘ , 2000‘ :r , ,5 ,1; .5: 1‘ I a. 15.00 I 2 ’4 I i IE I 5.00 « Willi .32? . fl ’. '. Maximum Stress at Failure, o.oo- Unagedi Aged Unagedl Aged Manufacturer I 1 Manufacturer l Manufacturer Il Manufacturer II (Uncoated) i (Coated) (Uncoated) (Coated) Bar Type I Pullout Fa‘lure DConcrete Failure I Bar Failure Figure 4.12. Bond Strength Test Results for Uncoated and Coated GFRP Bars in Unaged and Aged Condition. Pullout of GFRP bars from concrete was the predominant failure mechanism for unaged Specimens. Splitting of concrete occurred in most aged specimens in both uncoated and coated conditions (Figure 4.13). Unaged and aged concrete specimens 62 developed similar compressive strengths of 26.7 and 25.6 MPa (3,876 psi and 3,709 psi), respectively. This suggests that aging changed the bond failure mode because it produced some interaction between GFRP rebars and concrete. Our hypothesis is that, due to the thermal expansion mismatch of GFRP bars (in radial direction) and concrete, tensile stresses develop in concrete at the elevated temperature of aging (70 °C [158 °F]). These tensile stresses essentially produce microcracks in concrete, which lower the tensile strength of concrete and thus promote the split-cracking mode of bond failure (in concrete) in lieu of bar pullout. The observations reported in the literature support this hypothesis.4345 The details of pullout specimen would determine if concrete microcracking could lower pullout strength of GFRP bars. One should not neglect the potential for adverse effects of GP RP bar degradation on the alkaline environment of concrete on their pullout strength. 63 re, MPa Maximum Stress at Failu N O) Figure 4.13. Bond Strength Values for Different Failure Mechanisms. Coated; U = Uncoated), 64 ‘ E J, 29 m o 6'1 28 E 25 — - 2- g o g 27 % 24L ’6 26 “- Ll. "‘ ‘5 3 23 - a 25 8 2 5 22 o 03 24 E o a 2 E E 3 '§ 21 — i g ., s 22 20 1 TA 1 "‘ 21 "C W ll—U Bar Group (a) Pullout Failure 29 , I I O 28 — 27 o o 26 Aging Conditions 0 Aged O Unaged 25 —i I 1 HS "-0 Il-U Bar Group (C) Bar Failure 1 T 1 1 "C "U “-C II-U Bar Group (1)) Concrete Failure (c (a) Pullout Failure (b) Concrete Failure (c) Bar Failure Figure 4.14. Bond Strength Failure Mechanisms. 65 CHAPTER 5 CONCLUSIONS GFRP bars were subjected to accelerated alkali attack in order to simulate the long-term effects of concrete pore solution on GFRP reinforcement. The effect of fiber volume fraction, bar diameter, epoxy coating, aging method (wet-dry vs. continuous immersion), aging period, and matrix composition were investigated. Tensile strength tests were performed in order to assess the percent loss in tensile strength associated with alkali attack. The effect of accelerated aging on bond strength of GFRP bars to concrete was also investigated. Uncoated and coated bars embedded in concrete were placed in a heated water tank for six weeks. Bond tests were conducted to determine the effects of concrete alkaline environment on bond strength of coated and uncoated GF RP bars. The following conclusions were drawn from the experimental results: 1. Alkali attack caused significant loss of tensile strength in glass fiber composite bars. The loss in tensile strength under alkali attack increased with time Up to a certain age; different accelerated aging processes (wet-dry vs. continuous immersion) produced similar effects on GFRP bars. 2. Fiber volume fraction had a significant effect on alkali resistance of GFRP bars; bars with lower fiber volume fiactions (Vr = 50%) suffered less damage from exposure to highly alkaline environments than bars with higher fiber volume fractions (V r 2 60%). This is due to the higher amount of matrix material surrounding fibers, providing additional protection against the alkaline environment. Even though lower fiber volume fractions in GFRP bars entail lower initial tensile strengths, their higher 66 alkali resistance allows for the preservation of tensile strength over time, thus yielding economic benefits. Further investigations for verification of this conclusion are needed. Although analysis of the effect of bar diameter on alkali resistance was inconclusive, it is believed that GFRP bars With larger diameters could provide better protection against alkali attack. Increased radial distance of a greater fraction of fibers from the surface of larger bars should make it more difficult for detrimental ions from concrete pore solution to attack the bulk of fibers in larger diameter bars. . Epoxy coating of GFRP bars generally improved their alkali resistance. The trends were, however, inconsistent; more investigations are needed to further clarify the coating effects on alkali resistance of GP RP bars. . The results concerning matrix type effects on alkali resistance of GFRP bars were inconclusive. The effects of fiber volume fraction could have potentially overshadowed any adverse effects of lowerfcost matrix systems. This requires further investigation. For the specific geometry of bond test specimens considered in this investigation, accelerated aging did not significantly alter the peak bond strength, but altered the failure mechanism of the bond samples. The predominance of concrete splitting failure afier aging pointed at damage to concrete caused by aging effects, which could be explained by the thermal expansion mismatch of GFRP bars and concrete. 67 APPENDICES 68 APPENDIX A RAW TENSION TEST DATA Table A.1. Tensile Strength Results of GFRP Bars from Manufacturer 1. (a) Uncoated GFRP Bars Sample ID. Duration of Aging A9109 Method Bar Diameterl Tensile Strength (days) (inches) (MPa) ACTR-01 0 Control 0.25 72946 ACTR-02 0 Control I 0.25 80903 ACTR-03 0 Control [ 0.25 72506 ACTR-O4 0 Control I 0.25 31 7,64 ACTR-05 0 Control 0.25 31 3,49 ACTR-06 0 Control 0.25 823.35 ACTR-O7 0 Control 0 .25 819.98 ACTR-08 0 Control 0.25 820.49 ‘ ACTR—09 0 Control 0.25 718.05 ‘ ACTR-1 0 0 Control 0.25 724.54‘ ACS14-01 1 4 Immersion 0.25 645.96 ACS14—02 1 4 Immersion 0.25 608.61 ACS14-El 1 4 Immersion 0.25 533_12 [ACS1 4-04 \ 1 4 Immersion 0.25 507.21 {31081 4-05 \ 14 Immersion 0.25 456.64 ACS14-06 14 Immersion 0.25 597.98 ACS14—07 14 Immersion 0.25 409.95 ACS14-08 14 Immersion 0.25 589.68 ACS14-09 14 Immersion 0.25 394.39 ACSI4-10 14 Immersion 0.25 492.17 AWD1 4-01 14 Wet-dry 0.25 507.21 AWD14-02 14 Wet-dry 0.25 468.05 AWD14-03 14 Wet-dry 0.25 541.96 AWD14-04 14 Wet-dry 0.25 472.46 Awo14-05 14’ Wet-m 0.25 629.36 AWD14-06 14 Wet-dry 0.25 635.84 AWD14-07 14 Wet-dry 0.25 616.13 AWD14-08 1 4 Wet-dry 0.25 440.56 a AWD1 4—09 14 Wet-dry 0.25 676.56 AWD14-10 1 4 Wet-dry 0.25 675.00 . 69 Table A.1(a) (cont’d). Sample ID. Duration of Aging Aging Method Bar Diameter Tensile Strength (days) (inches) (MPa) ACSZ8-01 28 Immersion 0.25 378.83 ACSZ8—02 23 Immersion l 0.25 357.31 AC628-03 28 Immersion l 0.25 553.37 ACS28-04 28 Immersion 0.25 I 374,42 AC528-05 28 Immersion 0.25 57645 AC828-06 28 Immersion 0.25 45249 ACS28-O7 28 Immersion 0.25 575.16 Acsz8-08 28 lmmerslon i 0.25 52121 AC828-09 28 Immersion l 0.25 53320 AC828-10 28 Immersion I 0.25 42448 AW028-01 28 Wet-dry 0.25 516.54 AW028-02 28 Wet-dry 0.25 41255 AWDZ8-03 28 Wet-dry 0.25 331 .95 AW028-04 28 Wet-dry 0.25 436.41 AWD28-O5 28 Wet-dry 0.25 522.51 “ AWD28-06 28 Wet-dry 0.25 564.52 T“ AWD28-O7 28 Wet-dry 0.25 370.02fi‘ AW028-08 28 Wet-dry 0.25 51 5. 25 AWD28-09 28 Wet-dry 0.25 426.03 AW028-10 28 Wet-dry 0.25 471.68 RCSQ-OTl 42 Immersion 0.25 51 1 _33 Rosa-02 42 Immersion 0.25 353.68 ACS42-03 42 Immersion 0.25 276.65 ACS42-04 42 Immersion 0.25 592.01 ACS42-05 42 Immersion 0.25 538.07 ACS42-06 42 Immersion 0.25 596.94 ACS42-07 42 Immersion 0.25 317.63 ACS42-08 42 Immersion 0.25 507.99 ACS42-09 42 Immersion 0.25 320.74 ACS42-10 42 Immersion 0.25 568.41 AWD42-01 42 Wet-(EL 0.25 480.50 AWD42-02 42 Wet-dry 0.25 481 .01 AWD42-03 4L Wet-dry 0.25 534.96 AWD42-04 42' Wet-dry 0.25 343.56 AWD42-05 42 Wet-dry 0.25 390.76 a AWD42-06 42 Wet-dry 0.25 449.63 AWD42-07 42 Wet-dry 0.25 412.29 AWD42-08 42 Wet-dry 0.25 447.04 AWD42-09 42 Wet-dry 0.25 349.53 AWD42-1 0 42 Wet-dry 0.25 372.61 70 Table A.1(a) (cont’d). Sample ID. Duration of Aging Aging Method Bar Diameter Tensile Strength (days) (inches) jMPa) KCTR-01 0 Control 0.375 705.19 KCTR—02 0 Control I 0.375 638.86 KCTR-03 0 Control I 0.375 888.12 KCTR-04 0 Control I 0.375 636.71 KCS14-01 14 Immersion I 0.375 479.22 KCS14—02 14 Immersion 0.375 465.21 KCS14-03 14 Immersion 0.375 490.05 KCS14—04 14 Immersion 0.375 503.95 KCS14-05 14 Immersion 0.375 503.23 KCS14-06 14 Immersion I 0.375 54003 KCS14-07 14 Immersion I 0.375 49250 LKCS14-08 14 Immersion I 0.375 46593 I_KCS14-09 1 4 Immersion 0.375 50497 I KCS14-10 14 Immersion 0.375 479,73 KWD14-01 I 1 4 Wet-dry 0.375 520.71 KWD14-02 14 Wet-dry 0.375 488.21 ‘ KWD14—03 1 4 Wet-dry 0.375 520.81 ‘7 KWD14-04 1 4 Wet-dry 0375 520.30 ‘ KWDi4-05 14 Wet-dry 0.375 547.59 ‘ KWD14-06 1 4 Wet-dry 0.375 521_12 IK—WD14-07 14 Wet-dry 0.375 514.88 Won-E 14 Wet-dry 0.375 516—83 ‘ @01409 14 Wet-dry 0.375 526.33 Won-10 14 Wet—dry 0.375 500.37 Wcsza-m 28 immersion 0.375 468.89 KCSZ8-02 28 Immersion 0.375 499.45 KCSZ8-03 28 immersion 0.375 424.13 KCS28-04 28 Immersion 0.375 439.05 KCSZB-05 28 Immersion 0.375 471 .35 KCSZ8-06 28 Immersion 0.375 437.41 KCSZB-O? 28 Immersion 0.375 484.22 KCSZ8—08 28 Immersion 0.375 477.58 KC828-09 28’ Immersion 0.375 493.83 Kcsz8-10 28, Immersion 0.375 475.74 KWD28-01 28. Wet-dry 0.375 462.45 KWD28-02 28 Wet-dry 0.375 545.55 KWD28-03 28 Wet-dry 0.375 498.33 KWD28-04 28 Wet-dry 0.375 428.01 KWD28-05 28 Wet-dry 0.375 45§JJ,___ KWD28-O6 28 Wet-dry 0.375 511.41 Kw028-07 28 Wet-dry 0.375 488.52 Kw028-08 28 Wet-er 0.375 504.56 Kw028-09 28 Wet-dry 0.375 477.27 71 Table A.1(a) (cont’d). 72 Sample ID. Duration of Aging Aging Method Bar Diameter Tensile Strength (days) (inches) (MPa) Kw028-10 28 Wet-dry 0.375 500.17 KCS42-01 42 Immersion I 0.375 428.93 KCS42-02 42 Immersion I 0.375 442.42 KCS42-03 42 Immersion 0.375 423.41 KCS42-04 42 Im mersion 0.375 451 .72 KCS42-05 42 Immersion 0.375 43394 KCS42-06 42 Immersion 0.375 471.14 KCS42-07 42 Immersion 0.375 454.79 KCS42-08 42 Immersion 0.375 43046 KCS42-09 42 Immersion I 0.375 431.28 KCS42-10 42 Immersion I 0.375 43341 I KCS42-11 42 Immersion I 0.375 31957 IKCS42-12 42 Immersion 0.375 37691 I KWD42-01 42 Wetory 0.375 495.16 KWD42-02 42 Wet—dry 0.375 458.88 KWD42-03 42 Wet-dry 0.375 457.34‘ KWD42-04 42 Wet-dry 0.375 464.50‘ KWD42-05 42 Wet-dry 0.375 433.22‘ KWD42-06 42 Wet-dry 0.375 485.76 KWD42-07 42 Wet-dry 0.375 488.92 WWD42-O8 42 Wet-dry 0.375 458.47 Won-0M 42 Wet-dry 0.375 433—61 ‘ Won-1M 42 Wet-dry 0.375 51 5.19 (b) Coated GFRP Bars Sample ID. Duration of Aging Aging Method Bar Diameter Tensile Strength (days) (inchesL MPa) GCTR—01 0 Control 0.25 797.67 GCTR-02 0 Control 0.25 804.93 GCSZ8-01 28 Immersion 0.25 539.89 GCSZ8-02 28' Immersion 0.25 718.83 GCSZ8-03 2L Immersion 0.25 456.12 GCS28-04 2L Immersion 0.25 764.22 60828-05 28 Immersion 0.25 71 1 .05 60828-06 28 Immersion 0.25 539.63 GWD28-01 28 Wet-dry 0.25 787.04 Gw028-02 28 Wet-dry 0.25 619.24 GWD28—03 28 Wet-dry 0.25 539.1 1 GWD28-04 28 Wet-dry 0.25 658.j§______ GWD28-05 28 Wet-dry 0.25 398.54 GWD28-06 28 Wet-er 0.25 634k Sample ID. Duration of Agin Table A.l (b) (cont’d), ing Method Bar Diameter Tensile Str ength 9\A9 d3 (inches HCTR-OI 0 I Control 0.25 L 47:53 HCTR-02 0 I Control 0.25 304 :5 HCTR-03 0 Control 0.25 334'24 ions-04 0 k Control 0.25 729.21 Hcszam 28 ~ Immersion 0.25 466'23 £28432 !______2____8/_ Immel’SIOI'I 0.25 606.02 _IiCSZ8-O3 ‘___2_§_, I Immersion 0.25 537'09 HCS28—04 28 I Immersion I 0.25 549.74 Hcszsos 28 Immersion I 0.25 507:99 HCSZ8-06 28 Immersion 0.25 476.61 HWD28-01 28 Wet-dry 0.25 629,33 HWD28-02 Jr Wet-dry 0.25 702.23 l-iw028—03 ____,,,Z.§.— Wet-dry 0.25 769.40 HWD28-04 #232 Wet-dry 0.25 736.47 HWD28-05 28 Wet-dry 0.25 498.65 J HWDZ8-06 28 Wet-dry 0.25 575.67 A m 0 Control 0.375 682.91 m 0 Control 0.375 679.74 m 42 Immersion 0.375 343.08 m 42 Immersion 0.375 381.81 g m 42 Immersion 0.375 364.13 mm 0 Control 0.375 629.25 mm 0 Control 0.375 42 650.61 m 42 immersion 0.375; 400.82 m 42 Immersion 0.375 389.99 m 42 Immersion 0.375 Table A.2. Ten (a) Uncoated GFRP Bars sile Strength Results of GFRP Bars from Manufacturer 11. / Sample ID. Duration of Aging Aging Method Bar Diameter Tensile Strength gdaxg) (inches) (MPa) BCTR-01 0 Control 0.375 661 .00 BCTR-02 !_o’_’____ Control 0.375 655.42 BCTR-03 0 Jontrol 0.375 642.55 BCTR-04 #____Q_________fControI 0.375 629.44 BCTR-05 0 Control 0.375 648.02 BCTR-06 0 Control 0.375 677.88 BCTR-07 0 Control 0.375 649.71 BCTR—08 0 Control 0.375 649.84 73 Table A.2(a) (cont’d). ens/le '74 Table A.2(a) (cont’d), I n . . _ Sample ID Durat :a:;)Agan\Aging MethM\BZE;:meter Tensile Stren th es) 9 80542-01 42 rimmersion 0.375 I a BCS42-02 42 Immersion 0,375 BCS42-03 ____,_,4.2,.- Immersion 0.375 BCS42-04 42 Immersion 0.375 471 04 BCS42-05 42 k Immersion 0.375 471 '53 BCS42-06 42 4 I Immersion 0.375 474-81 80542-07 42 T Immersion 0.375 44568 80542-08 42 I ImmersionT (1375 472'50 BCS42—09 42 I Immersion I 0.375 454.90 BCS42-10 42 Immersion 0.375 43925 BCS42-1 1 42 Immersion 0.375 411.69 BCS42-12 42 Immersion 0.375 452.1 1 . 30542-13 42 Immersion 0.375 41443 / I Wet-dry 0.375 472.62 Wet-dry 0.375 495.68 BWD42-03 Wet-dry 0.375 528.94 J Wet-dry 0.375 473.23 A Wet-dry 0.375 485.49 42 Wet-dry 0.375 485.73 42 Wet-dry 0.375 499.69 42 Wet-dry 0.375 516.68 42 Wet-d 0.375 411.45 42 Wet-dry 0.375#_ 493.01 (b) Coated GFRP Bars Sample ID. Duration of Aging\Aging Method\Bar Diameter (days) (inches SCTR-01 0 Control 0.375 . 50542.01 42 I limon 0.375 #_ 575.07 50542.02 42 I immersion 0.375 521.17 TCT R-01 0 Control 0.375 697.1 7 0 Control 0.375 670.71 42 Immersion 0.375 461.70 42 Immersion 0.375 475.66 TCS42-03 42 Immersion 0.375 53792 75 Tabl e A.3- Tensile Strength Results of GFRP Bars from Manufa CtUrer III. (a) Uncoated GFRP Bars ensile 76 Table A.3(a) (cont’d). Sample ”1 Duration Of AQi“9\AQInQ Method\Bar Diameter Tensile S trength da 5 (inches 28 limes-mi 03.51 were CCS28-07 28 I Immersion 0.315 CCSZ8-08 28 Immersion 0.315 332.19 j 00528-09 28 Immersion 0,315 299 96 0052816 28 ~ immersion 0.315 349'1 7 CW828-01 28 A Wet-dry 0315 320:02 CWD28-02 28 I Wet-dry 0315 296.88 0w028-03 j Wet-dry I 0315 365.97 CW028-04 Wet-dry I 0_315 362.02 Wet-dry 0.315 391.63 Wet-dry 0.315 287.96 Wet-dry 0315 357.22 Wet-dry 0.315 329.11 Wet-dry 0.315 414.48 I Wet-dry 0.315 37454 (b) Coated GFRP Bars m 0 Control W 0.315 597.92 m 0 Control I 0.315 763.69 m 0 Control 0.31 5 f 692.38 mm 0 Control 0.315; 513.91 m 0 Control 0.315 635.83 0 Control Table A.4. Tensile Strength Results of GFRP Bars from Manufacturer IV. (a) Uncoated GFRP Bars Sample ID. Duration of AgingIAging Method Bar Diameter Tensile Strength (days) (inches) (MPa) NCTR-01 0 Control 0375 304.25 NCTR-02 0 WControl 0.375 302.92 NCTR-03 0 Control 0.375 358.67 NCTR-04 0 Control 0.375 379.1 1 NCTR-05 0 Control 0.375 318.27 NCTR-06 0 Control 0.375 385.64 NCTR-07 0 Control 0.375 380.19 NCTR-08 0 Control 0.375 363.99 NCTR-OQ 0 Control 0.375 343,91 77 Table A.4(a) (cont’d), on of Aging\Ag'ing Method Bar Diameter Tensile (inches ) Jmitrength Sample ID. Durati Control 0375 NCTR-10 0 38 0 I Control 0.375 43 7587: 0 Control 0.375 394'” 0 Control 0.375 503 .07 rNCTR-14 0 A Control 0.375 353.23 _I:lCTR—15 0 . Control 0.375 42567 NCTR-16 0 I Controi 0.375 513.47 N0528-01 I Immersion I 0.375 321 ‘90 N0528-02 Immersion I 0.375 397.25 NC828-03 28 Immersion 0.375 363.51 I NCS28—04 28 Immersion 0.375 424.34 NC328—05 28 Immersion 0.375 239.25 / Immersion 0.375 393.13 / Wet-dry 0.375 379.23 Wet-dry 0.375 443.89 NWD28—03 Wet-dry 0.375 366.29 _I Wet-dry 0.375 349.98 Wet-dry 0.375 349.23 28 Wet-dry 0.375 365.08 (b) Coated GFRP Bars Sample ID. Duration of Aging Aging MethodIBar Diameter Tensile Strength (Gaye) LC828-01 28 Immersion . LC828-02 28 Immersion LCS28-03 28 Immersion I LCSZB-04 28 I Immersion LC828-05 28 I Immersion LC828-06 28 I Immersion LW028-01 28 Wet-dry LW028-02 28 Wet—dry LWDZ8-03 28 Wet-dry LW028-04 28 Wet-dry LWDZ8-05 28 Wet-dry LWD28-06 28 Wet-dry MC828-01 28 Immersion MCSZ8—02 28 Immersion MCS28-03 28 Immersion M0828-04 28 Immersion MCSZB-OS 28 Immersion MCSZ8-06 28 Immersion 78 Sample ID. DU da 3 MWDZ8-O1 MW 028-02 MW028-03 MW 028-04 MW 028—05 MWDZB-OG Table A.4(b) (cont’d). ration of Aging\Aging Method\Bar Diameter T ‘ (inches) enSI/I: Strength 23 \ Wet-dry 0.375 pa 23 I Wet-dL 0.375 28 Wet-dry 0.375 /,2_§/‘ Wet-dry 0.375 456 99 2 Wet-dry 0.375 302 .43 23 Wet—dry 0.375 418.29 79 APPENDIX B RAW BOND TEST DATA Table 3.1. Bond Strength Results of GP RP Bars from Manufacturer I Sample ID. Coating Aging Condition Stres:MaFt) F)ailure Mode of Failure a KCTR-O1 Uncoated Una ed 26.88 Bar Pullout KCTR-02 Uncoated Una ed 20.02 Ba, Puuout KCTR-03 Uncoated Unaged 25.42 Bar Pullout KCTR-04 Uncoated Unaged I 21.75 Bar Pullout KCTR-05 Uncoated Unaged I 27.55 Bar Pullout KCS-O1 Uncoated Aged 22.01 Concrete Splitm KCS-02 Uncoated Aged 24.70 Concrete SplittifinEI UCTR-01 Coated Unaged 25.16 Bar Failure UCTR-02 Coated Unaged 28.42 Concrete Splitm UCTR-03 Coated Umd 29.32 Bar Pullout UCS-01 Coated Aged 25.16 Concrete Splim UCS-02 Coated Aged 25.16 Bar Pullout VCTR-01 Coated Ufiged 26.75 Concrete Splitting VCTR-02 Coated Uniaged 25.72 Concrete Splitting VCTR-03 Lgated UM 26.45 Concrete Sglitting ucs-01 Igated Aged 22.28 Concrete Silittin UCS-02 I Coated M 21.78 Concrete semi; Table B.2. Bond Strength Results of GFRP Bars from Manufacturer 11. Mode of Failure Manufacturer Sample l.D. Coating Aging Condition Stress at Failure (MPa) II BCTR-OI Uncoated um 38.83 Concrete Splitiin ll BCTR-02 Uncoated Unaggj 35.75 Concrete Splittln ll BCTR-03 Uncoated Unage_d 36.80 Bar Pullout ”7 II BCTR-04 Uncoated Unaged 30.83 Bar Pullout Ff II BCTR-05 Uncoated Unaggl 38.97 Bar Pullout '7 lI BCS-01 Uncoated mi 35.46 Concrete SplittirLg~ ll BCS-02 Uncoated Aged 41.16 Bar Failure 7 ii SCTR-01 anted Unaged 41.10 Bar Failure II SOS-01 Coated _Aged 36.10 Concrete Splittflg n TCTR-01 Coated Unaged 38.40 Concrete Spliflifi II TCTR-02 Coated Unaged 36.43 Concrete Splittfig W il TCTR—03 Coated Unaged 43,68 Ba, Fan”... I— ll TCS-01 Coated Aged 37.14 Concrete 5 Iittin Ifi II TCS-02 Coated Aged 36.23 Comm 8O REFERENCES 81 10. 11. REFERENCES . Altizer, S. D., Vijay, P. V., GangaRao, H. V. S., Douglass, N., Pauer, R., “Thermoset polymer performance under harsh environments to evaluate glass composite rebars for infrastructure applications”, Composites Institute ’3 51'" Annual Conference & Expo ’96, Feb 1996, pp. 3C1-3C19.(1) . Liao, K., Schultheisz, C. R., Huntson D. L., “Effects of Environmental Aging on the Properties of Pultruded GFRP”, Composites Part B: Engineering, v. 30 n. 5, 1999, pp. 485-493.(2) Tannous, F. E., Saadatmanesh, H., “Durability of AR glass fiber reinforced plastic bars”, Journal of Composites for Construction, v. 3 n. 1, Feb 1999, pp. 12-19.(3) Prian, L., Barkatt, A., “Degradation Mechanism of Fiber Reinforced Plastics and its Implications to Prediction of Long-Tenn Behavior”, Journal of Materials Science, v. 34 n. 16, 1999, pp. 3977-3989(4) Bank, L. C., Puterman, M.., “Microscopic Study of Surface Degradation of Glass Fiber-Reinforced Polymer Rods Embedded in Concrete Castings Subjected to Environmental Conditioning,” High Temperature and Environmental Eflects on Polymeric Coposites: 2nd vol., ASTM STP 1302, 1997, pp. 191-205(5) Coomarasamy, A., Goodman, 8., “Investigation of the Durability Characteristics of Fiber Reinforced Plastic (F RP) Materials in Concrete Environment”, Journal of Thermoplastic Composite Materials, v. 12 n. 3, May 1999, pp. 214-226(6) Pantuso, A., Spaeda, G., Swamy, R. N., “An experimental study on the durability of GFRP bars”, Fiber Composites in Infrastructure: vol. II. ICCI ’98, University of Arizona, Jan 1998, pp. 476-487.(7) Malvar, L.J., “Durability of Composites in Reinforced Concrete”, CDCC ’98-First International Conference on Durability of Composites for Construction, Sherbrooke (Quebec), Canada, Aug 1998(8) Sen, R., Mullins, G., Salem, T., “Durability of E-glassNinylester Bars in Alkali Solution,” Report to the US. Army and NFESC, University of South Florida, Tampa, FL, 1996(9) Tang, B., “Fiber Reinforced Polymer Composites Applications in USA”, First Korea/USA. Road Workshop Proceedings, Jan 1997.(10) American Concrete Institute: Committee 440, “Guide for the Design and Construction of Concrete Reinforced with FRP Bars”, Oct 2000, pp. 1-97.(11) 82 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Childs, P. J ., “Glass Rebar’s Growing Pains”, Civil Engineering, v. 69 n. 2, Feb 1999, pp. 65-67(12) Lincoln, J ., Liu, D., “Introduction to Composite Materials”, Materials Science and Mechanics Course Pack, Michigan State University, Jan 2001.(13) Bentur, A., Mindess, 8., “Fibre Reinforced Cementitious Composites”, Elsevier Applied Science Publishers Ltd., 1990.(14) Chawla, K. K., “Composite Materials Science and Engineering”, Springer-Verlag New York Inc., 1987.(15) Doremus, R. H., Glass Science, John Wiley New York, l973.(16) Angus, M. J., Glasser, F. P., “The Chemical Environment in Cement Matrices”, Materials Research Society Symposium Proceedings, v. 50, 1985.(17) Prian, L., Pollard, R., Shari, R., Mastropietro, C. W., Gentry, T. R., Bank, L. C., Barkatt, A., “Use of Thermogravimetn'c Analysis to Develop Accelerated Test Methods to Investigate Long-Term Environmental Effects on Fiber-Reinforced Plastics”, High Temperature and Environmental Eflects on Polymeric Composites: 2nd vol, ASTM STP 1302, 1997, pp. 206-222(18) Gentry, T. R., Bank, L. C., Barkatt, A., Prian, L., “ Accelerated Test Methods to Determine the Long-Term Behavior of Composite Highway Structures Subject to Environmental Loading”, Journal of Composites Technology and Research, JCTRER, v. 20 n. 1, Jan 1998, pp. 38-50.(19) Diefendorf, R. J ., “Carbon/Graphite Fibers”, Composites, Engineered Materials Handbook, ASM International.(20) Bank, L. C., Russell, T., “Accelerated Test Methods to Determine the Long-Term Behavior of FRP Composite Structures: Environmental Effects”, Journal of Reinforced Plastics and Composites, v. 14 n. 6, Jun 1995, pp. 559-587(21) Soroushian, P., Ravanbakhsh, S., Drzal, L. T., Non-Metallic Reinforcement of Concrete Bridge Decks: Final Report, Jan 2001 .(22) Shibley, A. M., Handbook of Composite Materials, Van Nostrand, New York, 1982, p. 448.(23) Tarnopol’skii, Yu. M., Bail’, A. 1., Fabrication of Composites, North Holland, Amsterdam, 1983, p. 45.(24) Slobodzinsky, A., Handbook of Composite Materials, Van Nostrand Reinhold, New York, 1982, p. 368.(25) 83 26. 27. 28. 29. 30. 31. 32. 33. 34. Lockwood, R. J ., Alberino, L. M., Advances in Urethane Science and Technology, 1981, v. 8.(26) Wu, W. P., “Thermomechanical Pr0perties of Fiber Reinforced Plastics (FRP) Bars,” Ph.D. diss., West Virginia University, Morgantown, West Virginia, 1990, pp. 292.(27) Faza, S.S., GangaRao, H. V. 8., “Theoretical and Experimental Correlation of Behavior of Concrete Beams Reinforced with Fiber Reinforced Plastic Rebars,” F iber—Reinforced Plastic Reinforcement for Concrete Structures, SP-138, American Concrete Institute, Farmington Hills, MI, 1993, pp. 599-614(28) Yamaguchi, T., Kato, Y., Nishimura, T., Uomoto, T., “Creep Rupture of FRP Rods Made of Aramid, Carbon and Glass Fibers,” Proceedings of the Third International Symposium on Non-Metallic (FRP) Reinforcement for Concrete Structures (FRPRCS- 3), Japan Concrete Institute, Sapporo, Japan, v. 2, 1997, pp. 179-186(29) Seki, H., Sekijima, K., Konno, T., “Test Method on Creep of Continuous Fiber Reinforcing Materials,” Proceedings of the Third International Symposium on Non- Metallic (F RP) Reinforcement for Concrete Structures WRPRCS—3), Sapporo, Japan, v. 2, 1997, pp. 195-202(30) Bank, L.C., “Properties of FRP Reinforcement for Concrete,” F iber—Reinforced- Plastic (F RP) Reinforcement for Concrete Structures: Properties and Applications, Developments in Civil Engineering, v. 42, A. Nanni, Ed., Elsevier, Amsterdam, 1993, pp. 59-86.(31) Al-Dulaijan, S.U., Nanni, A., Al-Zahrani, M.M., Bakis, C.E., “Bond Evaluation of Environmentally Conditioned GFRP/Concrete System,” Proceedings of the Second International Conference on Advanced Composite Materials in Bridges and Structures (A CMBS-Z), M. M. El-Badry, Ed., Canadian Society for Civil Engineering, Montreal, Quebec, 1996, pp. 845-852(32) Nanni, A., Nenninger, J., Ash, K., and Liu, J., “Experimental Bond Behavior of Hybrid Rods for Concrete Reinforcement,” Structural Engineering and Mechanics, v. 5, n. 4, 1997, pp. 339-354(33) Bakis, C. E., Al-Dulaijan, S. U., Nanni, A., Boothby, T E., Al-Zahrani, M. M., “Effect of Cyclic Loading on Bond Behavior of GFRP Rods Embedded in Concrete Beams,” Journal of Composites Technology and Research, v. 20, n. 1, 1998, pp. 29- 37134) 84 ICH III 3 AN IIIIIIIIIIIII 6