ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering L. C. Hollaway University of Surrey This chapter will introduce the polymers, fibres and advanced polymer composites used in bridge engineering, from the point of view of the basic components of composites, the manufacturing techniques of composites and the mechanical and in-service properties of the composite materials. Introduction The legacy of a lack of investment in research and development in the construction industry for more than 30 years following the end of the Second World War is clearly illustrated by the lack of progress in construction methods and in construction materials. Consequently, no interest was shown by potential material investors in civil engineering and the technological revolution in materials and their processing techniques utilised in other sectors of the manufacturing industry bypassed the construction industry. The Latham Report was published in 1994 and regarded the construction industry as low technology, low skilled and labour intensive compared with most other industries. It related the mismatch between research investment and construction expenditure to an inadequate understanding of many aspects of civil engineering, such as the deterioration mechanism for structures, and this has resulted in a lack of due allowance being made for practical repair and maintenance. However, notwithstanding this report, over the past 20 years there is evidence of a transition, by the construction industry, from the conventional civil engineering materials to the more advanced materials. One of the promising advanced materials to enter the civil engineering industry some 20 years ago was the advanced polymer composite (APC). The developments associated with the APC material, during this period, have been considerable and the requirements that have initiated this state have revolutionised the manufacturing and fabrication techniques for the production of the APC materials. The aerospace and defence industries have been utilising the APC materials for some considerable time and the manufacturing techniques which are currently used in construction have been developed from those used in the two industries. It will be demonstrated in this chapter that one area of the civil engineering industry which is realising great beneﬁts from the use of APCs is that of bridge engineering. Before discussing the subject of APC materials associated with bridge engineering it is essential for the reader to have doi: 10.1680/mobe.34525.0485 CONTENTS Introduction 485 Reinforcement mechanism of fibre-reinforced polymer composites 486 Advanced polymer composites 493 Adhesives 500 References 500 a clear understanding of the meaning of that material. Therefore, the deﬁnition which was adopted, in 1989, by the Study Group (on Advanced Polymer Composites) of the Institution of Structural Engineers, UK, will be given here. It was developed speciﬁcally for the construction industry from that produced by the British Plastics Federation for general polymer composites. The deﬁnition is as follows: Composite materials consist normally of two discrete phases, a continuous matrix which is often a resin, surrounding a ﬁbrous reinforcing structure. The reinforcement has high strength and stiﬀness whilst the matrix binds the ﬁbres together, allowing stress to be transferred from one ﬁbre to another producing consolidated structures. In advanced or high performance composites, high strength and stiﬀness ﬁbres are used in relatively high volume fractions whilst the orientation of the ﬁbres is controlled to enable high mechanical stresses to be carried safely. In the anisotropic nature of these materials lies their major advantage. The reinforcement can be tailored and orientated to follow the stress patterns in the component leading to much greater design economy than can be achieved with traditional isotropic materials. The reinforcements are typically glass, carbon or aramid ﬁbres in the form of continuous ﬁlament, tow or woven fabrics. The resins which confer distinctive properties such as heat or chemical resistance may be chosen from a wide spectrum of thermosetting or thermoplastic synthetic materials, but those commonly used are the thermoset resins, the epoxy, the phenolic and sometimes the polyester. More advanced heat resisting types such as bismaleimides are gaining useages in high performance applications and advanced carbon ﬁbre/thermoplastic composites are well into a market development phase. It will be seen later in this chapter that the main ﬁbres used in bridge engineering are the carbon, glass and aramid ﬁbres but these names are generic ﬁbre names and it is vitally necessary, when referring to these materials, to ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 485 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering deﬁne precisely which category of ﬁbre is being used/ described when discussing these materials. This will become clear in the section on Fibres where the various ﬁbres and their forms are discussed. Likewise, the manufacturing techniques for composites should be mentioned or their properties should be given when discussing/describing these materials as the various methods of production will produce diﬀerent forms of composite and therefore diﬀerent mechanical and in-service properties; this also will become clear in the section on Advanced polymer composites. The term ‘polymer composite material’ encompasses a wide range of ﬁbre–matrix materials each with their own unique characteristics. It is not the intention of this chapter to discuss the whole family of polymer composites; reference should be made to Hollaway (1993), Matthews and Rawlings (2000), and Hollaway and Head (2001) for further information on this topic. In this chapter the relevant advanced polymer composites and adhesive materials used in bridge management and their speciﬁc manufacturing techniques will be discussed, thus enabling the most eﬃcient and cost-eﬀective solution to the design problem to be made. It will then continue by describing the utilisation of the material in conjunction with the more conventional civil engineering materials used in bridge structures and as ‘all-composite’ bridge constructions. Currently there are no Eurocodes for the design of ﬁbre-reinforced polymer composites in civil engineering; these are being considered by the European Commission Joint Research Centre. There are UK design guides and speciﬁcations, European design guides, the ACI design codes USA and Canadian and Japanese design codes for APCs; these are given in the Appendix at the end of this chapter. Reinforcement mechanism of fibrereinforced polymer composites Introduction Advanced composite materials are usually referred to as ﬁbre-reinforced polymer composites (FRPs) and this term will now be used in all subsequent discussions. Fibrereinforced composite materials are made by the controlled distribution of two materials: these are (1) the continuous matrix phase (phase 1), (2) the ﬁbre reinforcement phase (phase 2). In addition there is the boundary between the matrix and the reinforcement (the interface area), which controls the properties of the given pair of materials. The continuous matrix phase (the polymer) The matrix phase (the polymer) is an organic material composed of molecules which are repeats of similar but simpler 486 www.icemanuals.com units called the monomer. The requirements of the matrix phase are (1) to bind and maintain the ﬁbres in position, (2) to protect the surfaces of the ﬁbres from external inﬂuences, environmental degradation and abrasion, and (3) to transfer stresses to the ﬁbres by adhesion and/or friction; the adhesion to the ﬁbres must be coupled with adequate matrix shear strength. In addition, the matrix must maintain chemical and thermal compatibility with the ﬁbre over the life span of the composite. Furthermore, during the manufacturing process complete wet-out of the ﬁbres by the matrix must be achieved. The three major types of matrices (polymers), which are used in bridge engineering, are the thermosetting, the thermoplastic and the elastomeric; each requires diﬀerent procedures for their manufacture. Their mechanical and in-service properties will be diﬀerent and these are discussed in sections on Thermosetting polymers and The mechanical and in-service properties of the thermosetting polymers. The three types of polymer are composed of long chain molecules and will be either (1) amorphous, which implies a random structure with a high concentration of molecular entanglement, or (2) crystalline, which has a high degree of molecular order or alignment. The random structure of the amorphous polymer, when heated, will become disentangled and the material will then be changed from a solid to a viscous liquid, whereas heating the crystalline polymer will change the material to an amorphous viscous liquid. It is noted that the matrix polymer material requires two components, namely the resin and curing agent (hardener). In bridge engineering there are two types of curing procedures: one uses a cold cure resin and the other uses a hot cure resin. The former would be used in situ in the ﬁeld to form either the composite and/or adhesive, when this latter is used. The cure temperature of the site would determine the length of the polymerisation period. It is advisable when using a cold cure resin to post cure the material at an elevated temperature. The length of time over which this cure takes place will depend upon the value of the elevated temperature; the higher the postcure temperature the lower will be the cure period. The hot cured resin products would be manufactured by automated techniques such as the pultrusion or the ﬁbre pre-impregnation in a factory under controlled conditions of temperature and pressure; the manufacturing and cure procedures are discussed in the section on The method of manufacture of the composite. Thermoplastic polymers Thermoplastic polymers are composed of long chain molecules which are held together by relatively weak Van der Waals forces but the chemical valency bond along the chains is extremely strong; they are generally of a semicrystalline form. Their strength and stiﬀness will be derived ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering from the very high molecular weight and from the inherent properties of the monomer units. The polyoleﬁns and polyesters are the main thermoplastics polymers which are utilised in bridge engineering for geotechnical applications. These materials are formed into geotextiles, geo-linear elements and geomembranes; geosynthetics are not discussed in this chapter but references can be found in Ingold and Miller (1988), Hollaway and Head (2001) and Fannin (2004). If the thermoplastic polymer, polyacrylonitrile (PAN), is placed under a tensile load the material behaves visco-elastically with a gradually decreasing slope of the stress–strain relationship. At a limiting load value the material will be unable to support further load and will continue to elongate excessively; the slope of the stress–strain relationship will become zero. During this elongation the molecules will align themselves with the direction of the applied load – the greater the load the greater the elongation – and in so doing a ﬁbre is formed. This ﬁbre is the precursor for the manufacture of carbon ﬁbres (see section on Carbon ﬁbres). Thermosetting polymers The thermosetting polymer consists of long chain molecules, which are cross-linked in a curing reaction producing covalent bonds between them; this chemical reaction is known as polycondensation, polymerisation or curing. This network, and the length and the density of the molecular units, are a function of the chemicals used in the manufacture of the polymers and the cross-linking is a function of the degree of cure of the polymer. Both the network and the cross-linking will have an inﬂuence on the mechanical and in-service properties of the material. Furthermore, the degree of cure is a function of the temperature and the length of the polymerisation period. The main thermosetting polymers used for structural components in bridge engineering are the epoxies, the vinylesters, and occasionally, the unsaturated polyesters. The elastomer A third member of the polymer family of interest to the bridge engineer is the elastomer, of similar form to that of rubber; the elastomeric material can be either reinforced or unreinforced. The material is composed of long chain molecules which are coiled and twisted in a random manner, similar in form to that of the thermoplastic polymer and due to the molecular ﬂexibility of the material it is able to undergo very large deformations. The molecules are cross-linked by a curing process known as vulcanisation; it is a similar procedure to the curing of the thermosetting polymer. The main use of this polymer in bridge engineering is for elastomeric bridge bearings; further discussions are given in the section on Elastomeric bridge bearings in the next chapter. Thermosetting polymers Epoxies The large family of epoxy resins represent some of the highest-performance resins of those available at this time. Epoxies generally out-perform most other resin types in terms of mechanical properties and resistance to environmental degradation; this has led to their almost exclusive use in aircraft components and a sizeable use in the construction industry. The most important epoxy resins for the bridge engineer are the low molecular weight polymers (oligomers), which are produced from the reaction of bisphenol A and epichlorohydrin. They range from the medium viscosity liquids through to high melting solids. Epoxies used in bridge engineering possess the following properties: n They have the capability of being used for diﬀerent composite manufacturing techniques, where polymerisation can take place at room temperature (cold cure resins) or at elevated temperatures (hot cure resins); thus two diﬀerent resin and curing systems (i.e. the cold or the hot cure systems) will be required for these manufacturing methods. Consideration must be given to the glass transition temperature (Tg ) of cold cure epoxy polymers (see the section on In-service properties). n They have high speciﬁc strengths and dimensional stability. n They have higher operating temperatures compared to that of the polyester due to their superior toughness. n They have low shrinkage during polymerisation and good adhesion to many substrates. This allows mouldings to be manufactured to a high quality and with good dimensional tolerance. n They use hot cured polymer systems which have a high temperature resistance and can be used at temperatures up to 2008C. Some epoxies have a maximum temperature range up to 3168C. Since the amine molecules ‘co-react’ with the epoxy molecules in a ﬁxed ratio, it is essential that the correct mix ratio is obtained between the epoxy resin and the curing agent to ensure that a complete reaction does take place. If amine and epoxy are not mixed in the correct ratios, unreacted resin or curing agent will remain within the matrix, which will aﬀect the ﬁnal properties after cure. To assist with the accurate mixing of the resin and curing agent (particularly for site fabrication), manufacturers usually formulate the two component parts to give a simple mix ratio that is readily achieved by measuring out the two components by weight or volume. Alternatively, the resin and curing agent are provided, by the manufacturers, in exact amounts in two separate containers; the resin container would be of suﬃcient volume to pour the curing agent into it and to mix by mechanical paddle. ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 487 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering Vinlyesters The vinylester is a hybrid form of polyester resin which has been toughened with epoxy molecules within the main molecular structure. Its molecule features fewer ester groups compared to the polyester resin and as these ester groups are susceptible to water degradation by hydrolysis, vinylesters exhibit better resistance to water and many other chemicals than do their polyester counterparts. Furthermore, as vinylesters have unsaturated esters of epoxy resins, they have similar mechanical and in-service properties to those of the epoxies, but they have similar processing techniques to those of the polyesters. Generally, the vinylesters have: n a processing method at both room temperature (cold cure resins) and elevated temperatures (hot cure resins); as for epoxy polymers, consideration must be given to the glass transition temperature (Tg ) of cold cure resins (see the section on In-service properties) n good wetting characteristics and bond well to glass ﬁbres n good resistance to strong acids and strong alkalis n reduced water absorption and shrinkage properties, as well as enhanced chemical resistance, compared to those of the polyester. Unsaturated polyesters Polyesters are not widely used for the manufacture of bridge structural components but they have been largely replaced by the vinylester polymers. However, they are used on occasions in the production of automated pultrusion components (see section on Automated processes). They are also readily processed and cured at ambient temperatures using a wet lay-up procedure (see section on Manual techniques); these ambient cured polymers must be post-cured. Styrene, in amounts of up to 50%, is the most commonly used diluent. However, it should be noted that during the wet lay-up (hand laminated) production method large amounts of styrene and trace quantities of styrene-7,8-oxide (SO) are released, which will cause a health risk unless safety precautions are taken. There are three types of commonly used polyester resins: the orthophthalic, the isophthalic and the bisphenol A. The ﬁrst resin mentioned has low thermal stability and chemical resistance. The isophthalic resins, which contain isophthalic acids as an essential ingredient, are of superior quality with better thermal and chemical properties. The bisphenol A resin is of a quality superior to both the orthophthalic and isophthalic resins with a higher degree of chemical, thermal and hardness resistance and has some degree ﬂame resistance. The shrinkage at cure of the polyester resin tends to be high, particularly if styrene in monomeric form is used as a reactive diluent in the resin. This will cause (1) an increase in hydrophobicity, thereby eﬀectively 488 www.icemanuals.com decreasing the level of moisture absorption, and (2) an increase in shrinkage up to levels of 5–19% by volume. This can result in signiﬁcant micro-cracking in resin-rich areas and high residual stresses in composites having high ﬁbre volume fractions. The polymerisation of thermosetting polymers must be completed before the polymer (or the FRP composite) is used in practice as the long-term performance properties of the composite will be inﬂuenced by the degree of cure of the resin, particularly for ambient (cold) cured systems. The section on the Method of manufacture of the composite discusses the various methods for the manufacture of composites and it will be seen that materials which are manually fabricated in the ﬁeld, together with adhesive polymers (if these are employed during fabrication), use cold cure systems at ambient temperatures. Composites, which are manufactured under factory conditions, use hot curing polymers and are cured at the operating temperature of the manufacturing process. The production temperature values of these processes are of the order of 120–1358C; the actual value will depend upon which system is being used. Ideally all cold cure adhesive polymers and polymer composites should be post cured at an elevated temperature for a certain length of time; the higher the temperature the shorter is the post-cure period. The initial cure of these polymers will reach only about 90% of the full polymerisation value after some ten days (depending upon the polymer system and ambient cure temperature). Cain et al. (2006) investigated the post-cure eﬀects on E-glass/vinyl-ester ﬁbre-reinforced polymer composites manufactured using the vacuum-assisted resin transfer moulding (VARTM) method (see the section on Method of manufacture of the composite). The composites were diﬀerentiated by varying levels of post-cure temperature and duration, and examined for the eﬀects of advancing cure at various points in the time after post-cure. In parallel, the matrix polymer was inspected using Fourier transform infrared spectroscopy to directly assess the degree of conversion. The results suggested the degree of conversion is limited to 80% for the vinylester oligomer and 90–95% for styrene following a post-cure of 938C. It was observed that after 300 days of ambient temperature storage, the non-post-cured samples approached the degree of conversion exhibited by those post-cured at 938C. The research team concluded that resin-dominated quasi-static properties are greatly aﬀected by the degree of cure whereas ﬁbre-dominated properties are not aﬀected so much. As mentioned above, this under-cure will aﬀect the long-term mechanical property of the polymer with the result that its exposure to a variety of adverse and sometimes harsh environmental conditions encountered in construction will degrade the polymer and the FRP composite material, and thus will alter their mechanical performance. A further concern regarding the ambient-cured systems is that their relatively low cure ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Advanced fibre polymer composite materials and their properties for bridge engineering temperature will aﬀect their glass transition temperature (Tg ) (see the section on In-service properties), which will generally be about 15–208C above the cure temperature. This point is discussed in the section on Dimensional stability. The mechanical and in-service properties of the thermosetting polymers The mechanical properties The important mechanical properties of the three polymers that are used in bridge engineering are given in Table 1. Although these mechanical properties are important to the bridge engineer, it should be remembered that the polymer composite derives its mechanical characteristics wholly from those of the ﬁbre. The most important properties of the polymer are its in-service properties as given in the section below. The in-service properties The most important in-service properties of polymers are listed below. The stiffness of thermosetting polymers As stated earlier, the stiﬀness is inﬂuenced by the degree of cross-linking of the polymer. The process of cross-linking produces a tightly bound three-dimensional network of polymer chains. This network, and the length and the density of the molecular units, are a function of the chemicals used in the manufacture of the polymer, and the cross-linking is a function of the degree of cure of the polymer. The strength of thermosetting polymers The short-term strength of a thermosetting polymer is dependent upon the bonding, length, density and the degree of cross-linking of the molecular structure of the polymer. It will also depend upon the type of loading applied to that polymer. The long-term strength of the polymer will be dependent upon its durability in the environment into which it is placed. The section on Moisture, aqueous and chemical solutions discusses the durability of Material Specific strength Ultimate tensile strength: MPa Modulus of elasticity in tension: GPa Coefficient of linear expansion: 10ÿ6 /8C Polyester 1.28 45–90 2.5–4.0 100–110 Vinylester 1.07 90 4.0 80 Epoxy 1.03 90–110 3.5 45–65 Table 1 Typical mechanical properties of the three thermosetting polymers used in bridge engineering ice | manuals composites which are largely dependent upon the durability of the polymer. Thermal conductivity of all polymers The thermal conductivity of polymers is low and consequently they are good heat insulators. The thermal conductivity of a polymer can be reduced by incorporating metallic ﬁllers at the time of manufacturing the polymer. Coefficient of thermal expansion of thermosetting polymers This characteristic is extremely important in design when considering joining FRP composites to a dissimilar material. The coeﬃcient of thermal expansion of a polymer is an order higher than that of the more conventional civil engineering materials. In bridge engineering practice the only time that this situation occurs is when an adhesive material is used to join two dissimilar materials, e.g. when FRP composite material is bonded to reinforced concrete, timber or to steel bridge structural members to strengthen/ stiﬀen them. This will be discussed in the sections on Adhesive bonding of polymer composites to steel adherends and Rehabilitating reinforced-concrete bridge engineering in the next chapter. Toughness of a thermosetting polymer The toughness of a brittle thermosetting polymer can be improved by blend ﬁlling or co-polymerising it with a tough but lower stiﬀness one. However, an increase in the toughness of the polymer will tend to decrease its stiﬀness and, consequently, a compromise between the strength and stiﬀness of a thermosetting polymer must be made. Permeability Polymers with high crystallinity/density or a high degree of cross-linking will generally have low permeability, thus gasses and other small particles will not readily permeate through it. Haque and Shamsuzzoha (2003), Liu et al. (2005) and Hackman and Hollaway (2006) have shown that the ingress of moisture will permeate through polymers over time, particularly if the polymer (composite) is permanently immersed in water or salt solution, or is exposed to de-icing salt solutions. Chemical resistance The chemical resistance of a thermosetting polymer depends upon the chemical composition and bonding in the monomer; generally, the resistance is good. Creep characteristics of polymers Polymers are classiﬁed as visco-elastic materials which indicate that they have both elastic solids and viscous ﬂuids characteristics. In assessing the creep of a polymer material it is important to know the following: ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 489 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering Stress Strain Stress Stress increase For 100 s duration 100 s 0.2 Log time Strain Log time (a) (b) (c) Note: For fibre–matrix composite materials As the fibre–matrix composite contains a visco-elastic polymer, this material will creep, however, the fibres do not creep and therefore stabilise the polymer thus causing the composite to creep less than that of the pure polymer but similar forms of creep curves to those above for polymers will be formed for composites. Figure 1 Creep curves: isochronous stress–strain curves (adapted from BS 4618 1970): (a) isostress creep curve for polymers (experimental/field values); (b) 100 s isochronous stress–strain curve for polymers (values derived from (a)); (c) isostrain creep curve for polymers (values derived from (a)) n the time-dependent nature of the micro-damage in the composite material subjected to stress. To minimise creep it is necessary to ensure that the service temperatures do not approach the glass transition temperature of the polymer. Figure 1 illustrates a family of creep curves consisting of isostress creep curve (Figure 1(a)), 100 s isochronous stress–strain creep curve (Figure 1(b)) and the isostrain creep curve (Figure 1(c)). These latter curves have been produced by cross-plotting, from the isostress creep curves at constant times. To produce isochronous stress–strain curves, BS 4618 (1970), ISO 899-1 (1993) speciﬁcation requires that the constant load tests are carried out under controlled conditions for the following durations 60 s, 100 s, 1 h, 2 h, 100 h, 1 year, 10 years and 80 years. This method yields a family of stress–strain curves, each relevant to a particular time of loading. From Figure 1(b), it will be seen that a family of creep curves, for any material, may be obtained by varying the stress. Isochronous stress–strain curves will correspond to a speciﬁc loading direction. Thus a 100 s isochronous stress–strain curve is a plot of the total strain (at the end of 100 s) against the corresponding stress level. The slope at any speciﬁc point on this curve will then deﬁne the creep modulus of the material at that particular stress level. The creep modulus will vary for every stress level. A sophisticated approach, to obtain creep characteristics of polymers, such as the time–temperature superposition principle (TTSP) has been developed by Aklonis and MacKnight (1983). By conducting a series of tests at diﬀerent temperatures (but the temperatures should not approach the Tg of the polymer), activation energy may be determined and through a kinetic approach of temperature–time 490 www.icemanuals.com The glass transition temperature of polymers The glass transition temperature (Tg ) of an amorphous material (the polymer) is the temperature at which it changes from (or to) a brittle or vitreous state to (or from) a plastic state; this change takes place over a temperature range and the mid-point of that range is taken as the Tg ; this is shown on Figure 2. The Tg depends upon the detailed chemical structure of the polymer which in turn depends upon the cross-linking of their molecules for strength. Crystalline polymers do not have a glass transition temperature but have a melting point, and possess some degree of amorphous structure; this portion usually makes up 40–70% of the polymer sample. Thus, a polymer can have both a glass transition temperature (the amorphous portion) and a melting temperature (the crystalline portion). The eﬀect of crystallinity on properties is observed Volume change n the loading histories and the nature of the applied load superposition plots, a master curve can be deduced and predictions made; this master curve has been discussed in Hollaway (1993). The time, applied stress superposition principle (TSSP) (Cessna, 1971) can also be used for polymers and polymer composite materials. Modulus of elasticity n the temperature and moisture environments into which it is placed Tg Temperature increase Tg Temperature increase Figure 2 Change in properties at the glass transition temperature of thermosetting polymers ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering when the polymer is at a temperature between the Tg and the crystalline melting point. All polymers below the Tg , whether crystalline or not, are rigid and frequently brittle; they therefore have both stiﬀness and strength. Above the Tg , the amorphous polymers are soft elastomers or viscous liquids, consequently they have no stiﬀness or strength, but the crystalline polymers will range in properties from a soft to a rigid solid depending upon the degree of crystallinity. The epoxies used in construction would generally be in the amorphous state, with a small amount of crystalline structure. The numerical value of the Tg may be quoted with slightly diﬀerent values depending upon the testing technique used; these methods are dynamic mechanical thermal analysis (DMTA) and diﬀerential scanning calorimetry (DSC). The drastic changes which any polymer will experience at the Tg is shown in Figure 2 as typical mechanical/physical properties. Modulus of elasticity (modulus of elasticity plotted against temperature) is shown in Figure 2(a); hardness (volume plotted against temperature) is shown in Figure 2(b). From the above discussion it will be clear that polymer composite structural members should not be exposed to temperatures above the Tg value of the matrix material. Hot cured thermosetting polymers would normally have a higher Tg value than the room temperature cured thermosetting materials; their actual values would generally be about 15–208C above the curing temperature. The Tg of some low-temperature (ambient cured) moulded composites can be increased in value by further post-curing the polymer at a higher temperature than that of the original cure but there is a maximum value of the Tg of all polymers irrespective of the post-cure temperature value. Dimensional stability The thermosetting polymers used in bridge engineering are semi-crystalline, semi-rigid and frequently brittle in nature and will lose their dimensional stability above the glass transition temperature (Tg ), therefore the strength and stiﬀness will be much reduced. If polymers with a high crystallinity are used they will have a region of acceptable dimensional stability above the Tg . The post-curing procedures of thermosetting polymers are vitally important, particularly the ambient cured ones as in this case the polymerisation will only be about 90% complete after about seven to ten days, depending upon the ambient temperature during the initial cure. The fibres The ﬁbres that are mainly used in bridge engineering for structural applications are carbon, glass and aramid ﬁbres and for geotechnical applications are those made from one of the polyoleﬁn family of thermoplastic polymers, (see Aboutaha, undated; Hollaway and Head, 2001; Chia-Nan Liu, 2003). Glass fibres Glass is the common name given to a number of mutually soluble oxides which can be supercooled without crystallising; they are all clear amorphous solids when cooled. Several grades of glass ﬁbre are produced commercially, of which the most important ones used in bridge engineering are as follows: n E-glass ﬁbre. This has a low alkali content of the order of 2%. It is used for general purpose structural applications and is the major ﬁbre used in the construction industry. It also has good heat and electrical resistance. n S-glass ﬁbre. This is a stronger (typically 40% greater strength at room temperature) and stiﬀer ﬁbre with a greater corrosion resistance than the E-glass ﬁbre. It has good heat resistance. The S-2-glass ﬁbre has the same glass composition as S-glass but diﬀers in its coating. The S-2-glass ﬁbre has good resistance to acids such as hydrochloric, nitric and sulphuric acids. n E-CR-glass ﬁbre. This has good resistance to acids and bases and has chemical stability in chemically corrosive environments. n R-glass ﬁbre. This ﬁbre has a higher tensile strength and tensile modulus and greater resistance to fatigue, aging and temperature corrosion than does E-glass. n Cemﬁl (or AR-glass) ﬁbre. This is speciﬁcally employed for resisting the alkali in Portland cement and is used as the reinforcement for glass-ﬁbre-reinforced cement (GFRC). Table 2 shows the chemical composition of E-glass and S-glass. The commercial manufacturing technique for glass ﬁbres is undertaken by drawing swiftly and continuously ﬁne ﬁlaments from the molten glass which is at a temperature of 1200–14008C. These series of ﬁlaments have exceptionally high speciﬁc stiﬀness and strength and range in diameter from 3 to 24 mm. During this production stage strands, each consisting of 200 individual ﬁlaments, are produced and a surface treatment or sizing is applied before the ﬁbres are gathered into strands and wound onto a drum. The manufacturing technique has been discussed in Hollaway and Head (2001). The mechanical properties of the various types of glass ﬁbre are given in Table 3. Carbon fibres The two high-performance carbon ﬁbres used in Europe, North America and Japan for bridge engineering are Oxides used CaO Al2 O3 For E-glass 17.5 14.0 For S-glass – 25 MgO B2 O 3 SiO2 4.5 10 54 – 65 10 Table 2 The chemical composition of E-glass and S-glass (after Phillips, 1989) ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 491 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering Material Fibre Glass fibre E A S-2 Carbon fibre PAN-based fibres: Hysol Grafil Apollo Elastic modulus: GPa 72.4 72.4 88.0 Tensile strength: MPa Ultimate strain: % 2400 3030 4600 3.5 3.5 5.7 IM HM† HS‡ 300 450 260 5200 3500 5020 1.73 0.77 2.00 BASF Celion G-40-700 Gy 80 300 572 4960 1860 1.66 0.33 Torayca T-300 234 3530 1.51 Pitch-based fibres: Hysol Union carbide T-300 T-400 T-500 T-600 T-700 227.5 – 241.3 241.3 248.2 2758.0 – 3447.5 4137.0 4550.7 1.76 – 1.79 1.80 1.81 Aramid fibre 49 29 125 83 2760 2750 2.40 4.00 High modulus (intermediate modulus); † ultra-high modulus (high modulus); ‡ high strain. Table 3 Typical mechanical properties of glass, carbon and aramid fibres (adapted from Hollaway and Head, 2001) deﬁned as the high modulus (H-M) carbon and the ultrahigh modulus (UH-M) ﬁbres, (in North America, Japan and some Asian countries these two deﬁnitions are referred to as intermediate-modulus and high-modulus carbon ﬁbres, respectively). The basic manufacturing techniques for the H-M and the UH-M ﬁbres are the same but the heat treatment temperature will be greater the higher the modulus of the ﬁbres, thus a more highly orientated ﬁbre of crystallites will be formed for the UH-M ﬁbre. Typical mechanical properties of carbon ﬁbres are given in Table 3. Hollaway and Head (2001) have stated that carbon ﬁbres are manufactured by controlled pyrolysis and crystallisation of certain organic precursors and the manufacturing process consists of sequences of procedures which include (1) stabilisation, (2) carbonisation and (3) graphitisation and ﬁnally (4) surface treatment; the manufacturing technique of the carbon ﬁbre is illustrated in Figure 3. The precursor ﬁbres that are used for the production of carbon ﬁbres, are (1) the polyacrylonitrile (PAN) and (2) the pitch ﬁbres. The ﬁrst ﬁbre precursor is manufactured by spinning to produce a round cross-section ﬁbre; the yield is only 50% of the original precursor ﬁbre. The PAN precursor carbon ﬁbres can also be manufactured by a melt-assisted extrusion as part of the spinning operation. I-type and X-type rectangular cross-section carbon ﬁbre composites are produced with a closer ﬁbre packing in the composite. The aerospace industry uses only these types of carbon ﬁbres and they are also mainly used in bridge engineering. The pitch precursor ﬁbres are derived from petroleum, asphalt, coal tar and PVC; the carbon yield is high but the uniformity of the ﬁbre cross-sections is not constant from batch to batch. This non-uniformity (a) Precursor (polyacrylonitrile) Either or T = 200–300°C (To achieve dimensional stability) T > 800°C (Carbon crystallites formed) T > 1200°C (Highly orientated fibres of crystallite formed) Oxygen is being absorbed Inert atmosphere Inert gas Stabilisation Carbonation Graphitisation (b) Pitch fibre Melt spinning (by-product obtained by the destructive distillation of coal) Pick-up spool Surface treatment Precursors to the carbon fibre used in civil engineering. (a) Polyacrylonitrile (PAN) fibre (used in the manufacture of high-modulus carbon fibres) (b) Pitch fibre produced from the distillation of coal (used for the manufacture of ultra-high-modulus carbon fibres) Figure 3 492 Manufacture of carbon fibre (adapted from Hollaway and Head, 2001) www.icemanuals.com ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering Aramid fibres Aramid ﬁbres are synthesised from the monomers 1,4phenyl-diamine ( para-phenylenediamine) and terephthaloyl chloride. The result is a polymeric aromatic amide (aramid) with alternating benzene rings and amide groups; these polymer strands, when produced, are randomly orientated. To manufacture aramid ﬁbres the polymer strands are dissolved and spun at a temperature of between ÿ508C and ÿ808C and are then extruded into a hot cylinder which is at a temperature of 2008C; this causes the solvent to evaporate and the resulting ﬁbre is wound onto a bobbin. The ﬁbre then undergoes a stretching and drawing process to increase its strength and stiﬀness properties. The resulting commercial ﬁbre has a high strength, is heat resistant and the value of its strength is unaﬀected by immersion in water. Two grades of stiﬀness are generally available: the ﬁrst has a modulus of elasticity of about 60 GPa and is used in applications such as cables and body armour; the second has a modulus of elasticity of 130 GPa and is used as reinforcement in polymer composites in construction. The aramid ﬁbre is an anisotropic material and, as is the case with all ﬁbres, gives higher strength and modulus values in its longitudinal direction compared with its transverse direction. The value of the longitudinal strength is 2.7 GPa and the ﬁbre is heat and cut resistant. It retains its mechanical properties from cryogenic temperatures up to 4008C and is resistant to fatigue, both static and dynamic, and is elastic in tension but behaves non-linearly Ultra-high-modulus carbon fibre High-modulus carbon fibre Aramid fibre Glass fibre Tensile stress: GPa is acceptable to the construction industry and the pitch ﬁbre precursor is invariably used when the ultra-high stiﬀness carbon ﬁbres are required. The carbon ﬁbres made from the pitch precursor tend to be cheaper than the PAN precursor. As the modulus of elasticity of the pitch ﬁbres is invariably much higher than that of the PAN carbon ﬁbres the strain to failure will be lower. Typically the diameter of carbon ﬁbre ﬁlaments are 5 and 8 mm and are combined into tows containing 5000–12 000 ﬁlaments. A common size of untwisted carbon ﬁbre ‘tow’ is called 12K which contains 12 000 ﬁlaments of carbon and is sold as (1) high-modulus ﬁbres (intermediate-modulus ﬁbres; stiﬀness 250–300 GPa), (2) ultra-high-modulus ﬁbres (high-modulus ﬁbres; stiﬀness 300–1000 GPa); the value of stiﬀness of the UH-M carbon ﬁbre used in bridge engineering is generally about 500 GPa. The tows can be twisted into yarns and woven into fabrics similar to that for glass ﬁbre. In addition to strong and high-stiﬀness ﬁbres, bridge engineering requires in-service resistance to withstand high temperature and aggressive environmental conditions; carbon ﬁbres, in general, are not aﬀected by moisture, solvents, bases and weak acids. Table 3 provides general and physical properties of epoxy polymers. Slopes of lines indicate modulus of elasticity 0 1 2 3 4 Tensile strain: % Figure 4 Typical comparative stress–strain characteristics of glass, aramid and carbon fibres in compression and in addition has a ductile compressive characteristic. The ﬁbre possesses good toughness and damage tolerance properties. It does not rust or corrode and its strength is not aﬀected by immersion in water. The main weakness of aramid ﬁbre is that it decomposes under alkaline or chlorine environments. Typical mechanical properties of aramid ﬁbres and their stiﬀness characteristics are given in Table 3 and Figure 4 respectively. There are various types of aramid ﬁbres on the market, including Kevlar1, which is made in the USA, Technora1, produced in Japan by Teijin, who also produce Twaron1. Advanced polymer composites The advanced polymer composites (APCs) are manufactured by the impregnation of the ﬁbre into the matrix material to form a composite system and in so doing to obtain a wide range of composites with varying mechanical properties. The ﬁbre–matrix composite can have the properties of anisotropy or isotropy by virtue of the arrangement and direction of the ﬁbres in the matrix. Mechanical properties of composites The mechanical properties of the ﬁnal ﬁbre–polymer composite will be dependent upon the following: n the method of manufacture n the mechanical properties of the component parts (i.e. carbon, glass or aramid ﬁbres and the polymer) n the relative proportions of the polymer and ﬁbre (ﬁbre volume fraction) n the orientation of the ﬁbre – that is, unidirectional (anisotropic composite), bidirectional aligned (orthotropic composite) or randomly orientated (quasi-isotropic composite) ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 493 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering n the long-term durability of the composite system; this will generally involve reduction in property values of the composite system from those of its pristine condition n the creep of the composite. Method of manufacture of the composite 2 The CFRP sheet cut to size is adhered to the concrete using the ﬁrst laminating layer of polymer; it is pressed down on the surface using a roller to expel entrapped air between the ﬁbres and the polymer. 3 After the paper backing is removed from the carbon ﬁbre sheet the second layer of impregnating polymer is applied. 4 Steps 2 and 3 are repeated in the case of multiple plies. 5 The system is then allowed to cure. There are a number of techniques available for the manufacture of advance polymer composites, all of which have an inﬂuence on the mechanical properties of the ﬁnal composite. As would be expected, the automated fabrication method tends to give higher values of strength and stiﬀness than the manual fabricated techniques because of the greater degree of process control, compaction and curing in the former techniques, thus providing a higher ﬁbre volume fraction. The various manufacturing processes given here are all relevant to bridge engineering and can be divided into three major divisions: (1) the manual production process, (2) the semi-automated process, and (3) the automated process. A short discussion of these processes and an indication when speciﬁc applications are used in practice now follows. The Dupont method is a system using Kevlar ﬁbres which is marketed as a repair system for concrete structures. The application of the material to the surface to be retroﬁtted is similar to the above. The Tonen Forca method is an unidirectional carbon ﬁbre sheet in an epoxy laminated system marketed in the UK by Kyokuto Boeki Kaisha Ltd. The system was originally developed by Mitsubishi Chemical Corporation and is therefore similar to the RePlark system. These methods have been used in bridge engineering to conﬁne columns or for strengthening/stiﬀening bridge beams (see the section on Rehabilitating reinforced-concrete bridge structures in the next chapter). Further details on these methods may be obtained from Hollaway and Head (2001). Manual techniques Semi-automated processes The manual processes used currently are a variation of the general wet lay-up method. The commercial companies which manufacture APCs by this process are: The semi-automated processes used currently are: n the REPLARK method n the resin transfer moulding (RTM) process n the Dupont method n the hot-melt factory-made prepreg. n the Tonen Forca method. The XXsys Technologies method was developed in the USA by XXsys Technologies, Inc., San Diego, California, as a semi-automated process for seismic retroﬁtting and strength restoration of concrete columns using continuous carbon ﬁbre. The XXsys carbon ﬁbre composite jackets are installed with a fully automated machine called RoboWrapper2 and portable oven for curing. The technology associated with the technique is based upon the ﬁlament winding of prepreg carbon ﬁbre tows around the structural unit thus forming a carbon ﬁbre jacket; currently, the structural unit to be upgraded would be a column. The polymer is then cured by a controlled elevated temperature oven and can, if desired, be coated with a resin to match the existing structure. An advantage of this automated process is that the carbon ﬁbre prepreg is impregnated with the polymer under factory-controlled conditions, providing good quality control and as a consequence a high strength-to-weight ratio. The equipment is erected on site with minimum disturbance to traﬃc and the whole operation is undertaken in minimum time; the latter will, however, depend upon the size of the job. The carbon ﬁbre jacket which is eventually formed around the column will increase the shear capacity of the column These techniques are basically the same with minor variations. The wet lay-up technique consists of in situ wetting of dry ﬁbres in the form of sheets or fabrics impregnated in situ with a polymer. The sheets or fabrics are wrapped around the structural member or placed on the mould and their size will depend upon the size of the member or mould, but ﬁbre reinforcements are generally of widths varying between 150 mm to 1500 mm. The REPLARK method was developed by the Mitsubishi Chemical Corporation, and is a pre-impregnated (prepreg) carbon ﬁbre sheet where the matrix material is an epoxy resin (Epotherm) and the ﬁbres are Mitsubishimanufactured ﬁbres, unidirectionally orientated. The sheet has a paper backing, which serves to keep the ﬁbres in position; the paper backing is removed when the sheet is placed in position on the mould structure. The procedure employed to apply the carbon-ﬁbre-reinforced polymer (CFRP) sheet is as follows: 1 The ﬁrst layer of impregnating polymer is applied to the prepared surface of the mould or structural member. 494 www.icemanuals.com n the resin infusion under ﬂexible tooling (RIFT) process ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Advanced fibre polymer composite materials and their properties for bridge engineering and will conﬁne the concrete and greatly enhance its ductility in the ﬂexural plastic hinge region. Furthermore, it will provide lap splice clamping and prevent local buckling of the vertical reinforcement. For corrosion-damaged columns, the jacket restores shear capacity and will prevent spalling of the cover concrete. Resin transfer moulding is a low-pressure, closed-mould semi-mechanised process. In RTM process, several layers of dry continuous strand mat, woven roving or cloth are placed in the bottom half of a two-part closed mould and a low-viscosity catalysed resin is injected under pressure into the mould cavity, and cured. Flat reinforcing layers, such as a continuous strand mat, or a ‘preform’ that has already been shaped to the desired product, can be used as the starting material in the RTM process. The potential advantages of RTM are the rapid manufacture of large, complex, high-performance structures with good surface ﬁnish on both sides, design ﬂexibility and the capability of integrating a large number of components into one part. This method can be employed to form large components for all composite bridge units but it is seldom used. The hot-melt factory-made pre-impregnated ﬁbre (prepreg) is cured in either the factory, for the production of precast plates, or on site if the prepreg composite is to be fabricated on to a bridge member. In the latter case a compatible ﬁlm adhesive is used and the adhesive and the prepreg components are cured in one operation under an elevated temperature of 658C applied for 16 h or 808C applied for 4 h; a vacuum-assisted pressure of 1 bar is applied for simultaneous compaction of the composite and the ﬁlm adhesive (see Hollaway et al., 2006). It is estimated that this method will be used increasingly for strengthening/stiffening degraded bridges (see the section on Rehabilitation of steel structural members in the next chapter). In the UK the manufacturing specialist in the production of hot-melt factory-made pre-impregnated ﬁbre for the construction industry is ACG in Derbyshire. Automated processes The automated processes which are available to the construction industry are: n the pultrusion technique n the ﬁlament winding technique. The pultrusion technique is used quite extensively in the construction industry and a large percentage of this use is associated with bridges. Sections are manufactured in a factory using a hot cure resin, the dies operating at temperatures between 1208C and 1358C. The pultrusion technique is associated with bridges (see the section on Rehabilitating reinforced concrete bridge structures). The sections produced will generally be fully cured but this does depend upon their size; the large sizes of the sections may require post-curing. ice | manuals Flat plates and various geometrical cross-sections can be produced; they are generally straight in the longitudinal direction although products can be manufactured which are curved in plan. Care must be taken to ensure that (1) the ﬁbres are well compacted into any bends in the crosssection (thus preventing voids forming), (2) there is complete wetting of the ﬁbres in the pultruded unit (again, preventing voids forming), and (3) the ﬁbres are well distributed in all cross-sections. It is not usual to have ﬁbre weight fractions (f.w.f.) greater than 60% although uncomplicated sections, such as small-diameter rods, have been manufactured with 70% f.w.f. To provide resistance to hostile environments, a resin-rich exterior surface to the pultruded section can be fabricated using a surface veil, which is incorporated into the structural component at the time of manufacture. The limiting factors on the size of the unit and the complexity of the cross-section is the pull force required to draw the pultruded section through the die; the more complex the section the greater will be the force due to friction of the unit within the die. Hydraulic pulling force systems up to 50 t on sections of 2.5 m 275 mm thick are currently in operation in the USA. Carbon, aramid and glass ﬁbres and epoxy, vinylester and polyester materials have all been used for the production of pultruded units. The epoxy polymers are probably the most diﬃcult to pultrude but do have low shrinkage during polymerisation (3–4%). The vinylester polymers have a shrinkage value of 6–10% and the polyester polymers have a large shrinkage during polymerisation of 12–19%. Further information may be obtained from Starr (2000) and Hollaway and Head (2001). Figure 5 illustrates the pultrusion technique. Some companies providing details of geometries and mechanical properties of pultruded proﬁles are Fiberline (Europe), Creative Putrusions, Strongwell and Bedford (USA). The pultrusion technique is utilised in bridge engineering to upgrade and retroﬁt structural beams, the manufacture of ‘all-composite’ bridges, bridge decks, near-surfacemounted rods, internal reinforcement to concrete and bridge enclosures and fairings. The ﬁlament winding technique is used to manufacture pressure pipes and to undertake wrapping of columns (see the section on FRP conﬁning of concrete in the next chapter). There are two diﬀerent winding methods: (1) wet winding, and (2) pre-impregnated winding, The wet winding method consists of continuous strands or roving of dry ﬁbres from a series of creels which are passed through a bath of ‘cold cure’ resin, cold curing agent, pigments and UV retardants on to a ‘pay-out eye’ which is mounted on a moving carriage along the length of a constant-speed rotating mandrel. The roving ﬁbre delivery system reciprocates along the length of the mandrel and is controlled relative to the rotation of the mandrel to give the required ﬁbre orientation. The speed of reciprocation and rotation ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 495 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering Heated die Puller Alternatively, fibres can be impregnated with resin by injection through portholes in the heated die as the fibres pass through Resin bath Creels of fibre (carbon and/or glass) Figure 5 A diagrammatic representation of the pultrusion machine (adapted from Hollaway and Head, 2001) are synchronised to hold a preset winding angle typically between 7–908. The machine has the ability to lay ﬁbres in any direction and to employ as many permutations of movements as is required by the structural design. After winding, the ﬁlament-wound mandrel is subjected to curing and post-curing operations during which the mandrel is continuously rotated to maintain uniformity of resin content around the circumference. After curing, the product is removed from the mandrel either by hydraulic or mechanical extraction. The pre-impregnated winding consists of passing pre-impregnated ﬁbres over a hot roller until tacky to the touch and they are then wound on to the rotating mandrel. Wet winding is the method usually used and has several advantages over pre-impregnated winding; these are low material cost, short winding time, and the resin formulation can be readily varied to meet speciﬁc requirements. Figure 6 shows a schematic representation of the ﬁlament winding technique. The XXSys Technology method, which is a site procedure for wrapping concrete bridge columns, is a ﬁlament winding technique; an example has been illustrated in Figure 7. Resin transfer moulding is a low-pressure, closed-mould semi-mechanised process. The process enables fabrication of simple low- to high-performance articles in varied sizes and proﬁles. The RTM process has been successfully used in moulding of complex three-dimensional shapes. In RTM, several layers of dry continuous strand mat, woven roving or cloth are placed in the bottom half of a twopart closed mould and a low-viscosity catalysed liquid resin is injected under pressure into the mould cavity, which is subsequently cured. Instead of using ﬂat reinforcing layers such as a continuous strand mat, the starting material in the RTM process can be a ‘preform’ that already has the shape of the desired product. The potential advantages of RTM can be summarised as rapid manufacture of complex, high-performance structures with good Traversing resin bath Resin bath Rovings Gear box Drive motor Rotating mandrel Figure 6 Diagrammatic representation of a filament winding machine (adapted from Hollaway and Head, 2001) 496 www.icemanuals.com Figure 7 Column wrapping using XXsys technique (by kind permission of XXsys Technologies, Inc., San Diego, CA, USA) ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering E-glass Specific weight 1.90 Tensile strength: MPa Tensile modulus: GPa Flexural strength: MPa Flexural modulus: GPa 760–1030 41.0 1448 41.0 S-2 glass 1.80 1690.0 52.0 – – Aramid 58 1.45 1150–1380 70–107 – – Carbon (PAN) 1.60 2689–1930 130–172 1593 110.0 Carbon (Pitch) 1.80 1380–1480 331–440 – – Table 4 Typical tensile mechanical properties of long directionally aligned fibre-reinforced composites (fibre weight fraction 65%) manufactured by automated process (the matrix material is epoxy) (adapted from Hollaway and Head, 2001) surface ﬁnish on both sides, design ﬂexibility and capability of integrating a large number of components into one part. The technique is sometimes used to manufacture bridge street furniture. There are three main industrial methods, all of which are a variation of the general RTM method; these are the Seemann composites resin infusion manufacturing (SCRIMP) method, the VA-RTM (vacuum-assisted RTM) and the TERTM2 method. Randomly orientated laminate 90° 75° Cross-ply laminate 60° Strength or stiffness Material 45° 30° 15° Strength or stiffness Unidirectional laminate Minor axis of laminate 90° Plane of strengh 0° Major axis of laminate Figure 8 The relationship between the stiffness–strength at angle to the major axes (08 direction) for the different types of polymer–fibre composites latter being equivalent to the matrix stiﬀness. Conversely, the randomly orientated ﬁbre composite gives equal properties in all directions from 08 to 908 axes direction. The long-term durability of the composite The civil engineering environments which cause the major concern for the durability of FRP composites are: Mechanical properties of the component parts of the composite n moisture and aqueous solutions (alkaline environments) The properties of the composite are highly dependent upon the type of ﬁbre used. The three ﬁbres used in bridge engineering, namely carbon, aramid and glass ﬁbre, all have diﬀerent stiﬀness values and the higher the stiﬀness of the ﬁbre (Table 3) the greater will be the stiﬀness of the composite; the orientation of the ﬁbre also plays a large part in the stiﬀness of the composite (see the section on the Orientation of the ﬁbre). n thermal eﬀects The relative proportions of the fibre and matrix Moisture, aqueous and chemical solutions The ﬁbre is the load-carrying component of the composite material, therefore the greater the ﬁbre volume fraction the stronger the composite. The orientation of the fibre The direction of the ﬁbres – that is, either unidirectional (anisotropic composite properties) or bidirectional (orthotropic (i.e. special type of anisotropy) composite properties) angle ply or randomly orientated ﬁbres (quasi-isotropic) – will determine the strength and stiﬀness of the composite. Figure 8 shows the relationship between the stiﬀness at angle to the major axes (08 direction) for the diﬀerent types of polymer–ﬁbre composites. It will be observed that for the unidirectional ﬁbre composite the major axes (08 direction) of the composite give the greatest stiﬀness and the 908 axes direction gives the least stiﬀness, this n ﬁre n ultraviolet radiation. Furthermore, the long-term loading conditions in which the material may have to function are: n fatigue n creep. Durability of FRP composites is one area which does require further research in order to gain a greater understanding of the characteristics of the material over long periods of time. All engineering materials are sensitive to environmental changes in diﬀerent ways which may lead to their degradation over time. FRP materials are no exception to this rule but they do oﬀer some signiﬁcant durability advantages over the more conventional construction materials. One such advantage is associated with their use as external reinforcement for reinforced concrete (RC) or metallic bridge members as they have good resistance to corrosion. A further example is the use of FRP rebars in RC members. One of the problems in acquiring data relating to the durability properties of any material is the length of time involved in gathering the relevant information. This is particularly diﬃcult with respect to construction polymer composites as there are many diﬀerent ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 497 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering polymers on the market and some of these have been modiﬁed by chemists over the years to improve their performance and as a consequence it would be hoped their durability performance as well. Moreover, modern structural FRP composites have been used in bridge engineering construction for some 15 years only. The majority of those that are used will have additives incorporated into them to reduce the length of cure time or to improve some speciﬁc mechanical or physical property. In eﬀect, the addition of additives to composites is equivalent to adding impurities to the polymer; these additives tend to reduce their mechanical properties. Consequently, many of the FRP composite bridge components that are being, or have recently been, erected are monitored for any changes in their stiﬀness characteristics or any signs of degradation. Indeed, bridge structures are visually inspected at regular intervals and any degradation which has occurred will be reported and rectiﬁed accordingly. To obtain the durability characteristics of a composite as quickly as possible, sometimes test specimens are exposed to an accelerated test regime which generally involves the specimens being subjected to an environment many times more severe than that which would be experienced in practice. In addition, these test samples are exposed to an elevated temperature to further increase degradation. This method of accelerating the degradation of polymer composites should be used with caution. High temperatures are not relevant for most civil engineering polymer composites that operate at normal environmental conditions; furthermore, the degradation mechanisms in FRP materials at high temperatures are diﬀerent to those under the lower practical temperatures. It will also be realised that as the environmental temperature rises towards the glass transition temperature of the polymer, the latter will lose some of its stiﬀness and strength with the result that the investigation will not be analysing the original material. The accelerated test procedures for obtaining durability data under one environmental situation will generally not be equivalent or relevant to the more gradual degradation eﬀect had the environment been applied in a less rigorous manner. Moreover, materials used in construction would normally be exposed to many diﬀerent environments acting simultaneously but in a less harsh way, and each environment possibly having an eﬀect upon the other. Monitored ﬁeld tests are the most relevant tests to establish the durability characteristics of polymer composites used in bridge construction but the disadvantage of this method of obtaining information is the length of time involved. Hollaway (2007) (and see also Hollaway (2008a, b) where the following discussion has been reported due to the importance the author attaches to the understanding and interpretation of the accelerated test results) has discussed many ﬁeld surveys of bridges and other structures which have been and are being undertaken throughout 498 www.icemanuals.com the world. The results from these tests have revealed or are revealing interesting discoveries regarding the resistance of FRP composites to the natural or speciﬁc environments. One interesting example is associated with the degradation of glass-ﬁbre-reinforced polymer (GFRP) rebars in concrete. Accelerated laboratory test results of GFRP in a simulated concrete pore water solution of high pH values and at elevated temperatures up to 808C have indicated that there is a decrease in the tensile, shear and bond strengths for that material (Bank and Gentry, 1995; Bank et al., 1998; Sen et al., 2002). Uomoto (2000) has concluded that these results would suggest that there is a case for not using GFRP rebars in concrete. However, Tomosawa and Nakatsuji (1997) have shown that after 12 months’ exposure to alkaline solutions at a temperature between 20 and 308C, and Clarke and Sheard (1998), likewise, after two years’ exposure to a tropical climate on a test platform oﬀ the Japanese coast, have reported that there had been no material or physical deterioration to the GFRP composite. Furthermore, Sheard et al. (1997) reported that the overall conclusions of the work of the EUROCRETE project were that GFRP is suitable in a concrete environment. In the section on Internal reinforcement to concrete members in the next chapter, a discussion is given of the tests undertaken by ISIS, Canada Research Network of Centres of Excellence, on the reliability of ﬁve GFRP RC structures to provide information on the reliability of GFRP materials. Methods to improve the durability of FRP composite materials in the civil infrastructure have been discussed in Hollaway (2007), and will not be repeated here. Fire behaviour of polymer composites A major concern of the engineer using polymers in the construction industry is the problem associated with ﬁre. The polymer component of the composite is an organic material and is composed of carbon, hydrogen and nitrogen atoms; they are ﬂammable to varying degrees. It is possible, however, to incorporate additives into the resin formulations, to add nano-clay particles to the pristine polymer, Hackman and Hollaway (2006), or to alter the chemical structure of the polymer, thereby modifying the burning behaviour and producing a composite with a much enhanced ﬁre property. Nevertheless, ﬁre can damage composites and indeed all civil engineering materials. Many composites used in bridge engineering have high ﬁbre volume fractions therefore the progress of the ﬁre is reduced considerably as the ﬁbre does not burn and the penetration progress of the ﬁre into the interior of the composite through the burning of the polymer is much reduced. FRP composites are widely used in the rehabilitation of bridge beams (see sections on Adhesive bonding of polymer composites to steel adherends to Seismic retroﬁt of RC structures in the next chapter) and bridge columns (see section on FRP conﬁning of concrete in the next chapter) ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering which are constructed from either reinforced concrete or metallic materials. The potentially harmful eﬀects of ﬁre and high temperature on FRP-strengthened reinforced concrete and metallic bridge structures are well recognised in the literature, although few studies have been undertaken on this topic. The ﬁrst researchers to investigate FRP composites bonded to RC beams were Deuring (1994) and Blontrock et al. (2001) and limited laboratory experimental work has been undertaken on wrapped cylindrical columns after exposure to elevated temperatures (see Saaﬁ and Romine, 2002). Bisby et al. (2004) undertook a numerical and experimental programme to investigate the ﬁre performance of FRP-wrapped fullscale reinforced-concrete columns in order to provide ﬁredesign guidance. The numerical modelling is capable of predicting the thermal and structural response of an FRPwrapped concrete column under exposure to a standard ﬁre. Thermal effects Coefficient of thermal expansion The glass, aramid and carbon ﬁbre components of ﬁbre– matrix composites have a much lower coeﬃcient of thermal expansion in the longitudinal direction than the matrix material and will be in the region of 10 10ÿ6 /8C, ÿ2 10ÿ6 /8C, and ÿ0:9 10ÿ6 /8C respectively, depending upon the type of speciﬁc ﬁbre; the value for the matrices are given in Table 1. The ﬁbres stabilise the composite system and reduce its coeﬃcient of thermal expansion to a value near to that of the conventional material; the actual value will depend upon the ﬁbre volume fraction and ﬁbre array. In addition, the value of the coeﬃcient of thermal expansion will vary with the environmental temperature range into which the composite is placed. Furthermore, the degree of cross-linking of the polymer will also inﬂuence the rate of thermal expansion. Thermal conductivity This property is particularly important when FRP composites are used in the production of FRP bridge decks (see the section on FRP bridge decks in the next chapter). The thermal conductivity of polymers is low and consequently they are good heat insulators (see the section on In-service properties). Thermal-stress analyses have been undertaken by Alnahhal et al. (2006), utilising the ﬁnite-element method to predict the failure mechanisms and the ‘ﬁre resistance limit’ of a superstructure of a bridge under extreme thermal loading conditions. The results were veriﬁed by undertaking ﬁeld results provided by the New York State Department of Transportation. In addition, damage simulations of the FRP deck as a result of snow and ice clearance by snowplough were performed to investigate any possibility of bridge failure after damage had occurred. Thermal simulations showed that FRP bridge decks are highly sensitive to the eﬀect of elevated temperatures. The FRP deck approached the ﬁre resistance limit at an early stage of the ﬁre incident under all cases of ﬁre scenarios. The damage simulations due to a snowplough showed minimal bridge failure under the worst-case damage scenario. These results provide an insight into the safety and reliability of the FRP systems after the stipulated damage scenarios are considered. The paper provides discussions concerning the recommended immediate actions necessary to repair the damaged region of the FRP deck panels. The creep of the composite The mechanics of creep in ﬁbre/polymer composite materials are related to: 1 the deformation characteristics of the creep of the viscoelastic matrix (see the section on Creep characteristics of polymers) 2 the near zero creep of the ﬁbres 3 the stabilising eﬀect of the ﬁbres on the polymer as a result of items 1 and 2 4 the progressive changes in the internal balance of forces within the materials resulting from the behaviour of the matrix, the ﬁbre, the adhesion and the load transfer at the ﬁbre–matrix interface. The deformational behaviour of an advanced polymer composite is a function of the resin type, the ﬁbre type and its architecture, ﬁbre volume fraction, direction of heat ﬂow, and service temperature. The mechanical behaviour of composite materials can be established by applying constant loads over long periods of time; these investigations may be deﬁned as creep tests. The tests produce curves of elongation against time at diﬀerent stress levels and although they are not able to produce data that may be converted directly into stress–strain curves, constant time sections through families of such creep curves have been used to produce isochronous stress–strain curves (see the section on Creep characteristics of polymers). Tensile and compressive properties of polymer composites Tensile properties of composites Table 4 gives typical properties of composites manufactured using long directionally aligned ﬁbre reinforcement of glass, aramid and carbon ﬁbres with a ﬁbre/matrix ratio by weight of 65%. Typical tensile mechanical properties of glass ﬁbre composites manufactured by diﬀerent techniques are given in Table 5; the eﬀects that the methods of fabrication have on the properties is clearly shown. The variation of the composite tensile properties when the ﬁbre/ matrix ratio is changed but the method of manufacture ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 499 ice | manuals Advanced fibre polymer composite materials and their properties for bridge engineering Flexural strength: MPa Flexural modulus: GPa Tensile strength: MPa Tensile modulus: GPa Wet lay-up 62–344 4–31 110–550 Spray-up 35–124 6–12 83–190 5–9 RTM 138–193 3–10 207–310 8–15 Filament winding 550–1380 30–50 690–1725 34–48 Pultrusion 275–1240 21–41 517–14 448 21–41 Method of manufacture 6–28 Table 5 Typical tensile mechanical properties of glass fibre composites manufactured by different fabrication methods (adapted from Hollaway and Head, 2001) and component parts of the composite remain constant is shown in Table 6. Composites under compressive loads The integrity of both component parts of the composite is far more important under a compressive load than under a tensile load. Furthermore, local resin and interface damage caused by compressive loading leads to ﬁbre instability which is more severe than the ﬁbre isolation mode which occurs in tensile loading. The mode of failure for FRP composites subjected to longitudinal compression will depend upon the type of ﬁbre, the ﬁbre volume fraction, the type of resin and may include ﬁbre micro-buckling, transverse tensile failure or shear failure. The compressive strengths of CFRP and GFRP composite materials increase as the tensile strengths increase, but the aramid ﬁbres in aramid-ﬁbre-reinforced polymer (AFRP) composites exhibit non-linear behaviour in compression at a relatively low level of stress. The value of the compressive modulus of elasticity of FRP materials is generally lower than their tensile value. Test samples containing 55–60% weight fraction of continuous ﬁbres in a vinyl-ester resin have compressive modulus of elasticity values of approximately 80%, 100% (but the compressive strength is very low) and 85% of the tensile value for GFRP, AFRP and CFRP respectively. Subramaniyan et al. (2003) have shown that by adding nanoclays to the polymer the compressive strengths of GFRP composites do increase. Flexural strength: MPa Flexural modulus: GPa Tensile strength: MPa Tensile modulus: GPa Fibre/matrix ratio: % Specific weight 67 1.84–1.90 483 17.9 269 19.3 65 1.75 406 15.1 214 15.8 50 1.80 332 15.3 166 15.8 Table 6 Typical tensile mechanical properties of glass fibre/vinylester polymer (compression moulding – randomly orientated fibres) (adapted from Hollaway and Head, 2001) 500 www.icemanuals.com Adhesives Structural adhesives used in bridge engineering are usually epoxy systems and can be (1) a two-part cold cure, (2) a two-part hot cure or (3) an adhesive ﬁlm; (2) and (3) require heat for polymerisation. A considerable amount of bonding, particularly when upgrading bridge structures, is required to be undertaken on site and therefore cold cure adhesives are generally used; the adhesive should be postcured to a temperature of at least 508C. The Tg (see the section on In-service properties) is then about 658C. If it is not post-cured the Tg value will only be about 158C above the ambient temperature of cure. If the bonding operation is undertaken in a factory environment the hot cure adhesive system is used and is cured under heat; the Tg is then dependent upon the heat of polymerisation. The cross-linked chemical structure of the adhesive renders the polymer insoluble and infusible and these characteristics greatly reduce the creep of the adhesive (see the section on Creep characteristics of polymers). To increase the toughness of an adhesive polymer it is possible to blend, ﬁll or co-polymerise it with a tough polymer. This will cause a reduction in strength of the adhesive and in this situation it is necessary to compromise between strength and stiﬀness. In the case of bonding of the FRP to metals, or when bonding FRP materials to the top surface of a bridge deck that is to receive hot bituminous surfacing, an adhesive with a high Tg must be used (The Concrete Society, 2000, 2003). Further information on adhesive bonding may be obtained from Mays and Hutchinson (1992). References Aboutaha R. S. (undated) Investigation of Durability of Wearing Surfaces for FRP Bridge Decks. Syracuse University and Cornell University, Project No. 01-50. Aklonis J. J. and MacKnight W. J. (1983). Introduction to polymer viscoelasticity 2nd edition. Wiley, New York, pp. 36–56. Alnahhal W. I., Chiewanichakorn M., Aref A. J. and Alampalli S. (2006) Temporal thermal behaviour and damage simulations of FRP deck. Journal of Bridge Engineering, 11, 4, 452–464. Bank L. C. and Gentry R. T. (1995) Accelerated test methods to determine the long-term behaviour of FRP composite structures: environmental eﬀects. Journal of Reinforced Plastics and Composites, 14, 559–587. Bank L. C., Gentry R. T., Barkatt A., Prian L., Wang F. and Mangla S. R. (1998) Accelerated aging of pultruded glass/ vinylester rods. Proceedings of the 2nd International Conference on Fibre Composites in Infrastructure (ICCI), 2, 423–437. Bisby L. A., Kodur V. K. R. and Green M. F. (2004) Performance in ﬁre of FRP-conﬁned reinforced concrete columns. Proceedings of the 4th International Conference on Advanced Composites Materials in Bridge and Structures, 20–23 July, Calgary, Alberta, Canada, 1–8. Blontrock H., Taerwe L. and Vandevelde P. (2001) Fire testing of concrete slabs strengthened with ﬁbre composite laminates. Proceedings of FRPRCS-5, London, UK, July, 547–556. ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Advanced fibre polymer composite materials and their properties for bridge engineering British Standards Institute: BS 4618-1.1 (1970). Recommendations for the presentation of plastics design data. Mechanical properties, Creep. BSI, London. British Standards Institute: BS ISO 899-1 (1993). Determination of creep–tensile creep. BSI, London. Cain J. J., Post N. J., Lesko J. J., Scott W., Case S. W., Lin Y.-N., Riﬄe J. S. and Hess P. E. (2006) Post-curing eﬀects on marine VARTM FRP composite material properties for test and implementation. Journal of Engineering Materials and Technology, 128, 1, 34–40. Cessna L. C. (1971) Stress–time superposition for creep data for polypropylene and coupled glass reinforced polypropylene. Polymer Engineering Science, 13, May, 211–219. Chia-Nan Liu (2003). Working-strain based design of land-ﬁll ﬁnal cover systems. Journal of the Chinese Institute of Engineers. 26, 2, 249–253 (in English). Clarke J. L. and Sheard P. (1998) Designing durable FRP reinforced concrete structures. Proceedings of the 1st International Conference on Durability of Fibre Reinforced Polymer (FRP) Composite for Construction (CDCC 1998), Sherbrooke, Quebec, Canada, 13–24. Deuring M. (1994) Brandversuche an nachtraglich verstärken trägern aus beton. Report No. 148795, Switzerland, EMPA (in German). Emmons P., Thomas J. and Sabnis G. M. (2001). New Strengthening Technology Developed – Blue Circle Cement Silo Repair and Upgrade. Proceedings of the Int. Workshop on Structural Composites for Infrastructure Applications. Cairo, Egypt, May 28–30, 2001, pp. 97–107. Fannin J. (2004) AASHTO M288. Durability considerations in Standard speciﬁcation documents. Proceedings of the 57th Canadian Geotechnical Conference, 5th Joint CGS/IAH–CNC Conference, Session 4D, 21–26. Hackman I. and Hollaway L. C. (2006) Epoxy-layered silicate nanocomposites in civil engineering. Composites Part A, 37, 8, pp. 1161–1170. Haque A. and Shamsuzzoha M. (2003) S2-glass/epoxy polymer nanocomposites manufacturing, structures, thermal and mechanical properties. Journal of Composite Materials, 37, 20, 1821–1837. Hollaway L. C. (1993) Polymer Composites for Civil and Structural Engineering. Blackie Academic and Professional, Glasgow. Hollaway L. C. (2007) Survey of ﬁeld applications. In: Durability of Composites for Civil Structural Applications (Karbhari V. M. (ed.)). Woodhead Publishing, Oxford, Ch. 12. Hollaway L. C. (2008a) Advanced Fibre Reinforced Polymer Composites of Advanced Composite Materials. World Scientiﬁc Publishing, Ch. 5. Hollaway L. C. (2008b) Manual of Construction Materials (Forde M. J. (ed.)). Thomas Telford, London, section 7. Hollaway L. C. and Head P. R. (2001) Advanced Polymer Composites and Polymers in the Civil Infrastructure. Elsevier, Oxford. Hollaway L. C., Zhang L., Photiou N. K., Teng J. G. and Zhang S. S. (2006) Advances in adhesive joining of carbon ﬁbre/polymer ice | manuals composites to steel members for repair and rehabilitation of bridge structures. Journal of Advances in Structural Engineering, 9, 6, 101–113. Ingold T. S. and Miller K. S. (1988) Geotextiles Handbook. Thomas Telford, London. Latham M. (1994) Constructing the Team. Final Report of the Government/Industry Review of Procurement and Contractual Arrangements in the UK Construction Industry. (The Latham Report) HMSO, London. Liu H. B., Zhao X. L. and Al-Mahaidi R. (2005) The eﬀect of fatigue loading on bond strength of CFRP bonded steel plate joints. Proceedings of the 1st International Symposium on Bond Behaviour of FRP in Structures (BBFS 2005) (Chen J. F. and Teng J. G. (eds)), International Institute for FRP in Construction, Hong Kong, pp. 459–464. Matthews F. L. and Rawlings R. D. (2000) Composite Materials: Engineering and Science. Woodhead Publishing, Cambridge, UK. Mays G. C. and Hutchinson A. R. (1992) Adhesives in Civil Engineering. Cambridge University Press, Cambridge, UK. Philips C. N. (1989) Design with Advanced Composite Materials. Published for The Design Council by Springer-Verlag, Berlin, Heidelberg, New York, London, Tokyo. Saaﬁ M. and Romine P. (2002) Eﬀects of ﬁre on concrete cylinders conﬁned with GFRP. Proceedings of Durability of FRP Composites for Construction Conference (CDCC ‘02), Montreal, PQ, Canada, 512–521. Sen R., Mullins G. and Salem T. (2002) Durability of E-glass/ vinylester reinforcement in alkaline solution. ACI Structural Journal, 99, 369–375. Sheard P., Clarke J. L., Dill M., Hammersley G. and Richardson D. (1997) EUROCRETE – taking account of durability for design of FRP reinforced concrete structures. Proceedings of the 3rd International Symposium on Non-metallic (FRP) Reinforcement for Concrete Structures, Sapporo, 2, 75–82. Starr T. F. (ed.) (2000) Pultrusion for Engineers. Woodhead Publishing, Cambridge, UK. Subramaniyan A. K., Bing Q., Nakaima D. and Sun C. T. (2003) Eﬀect of nanoclay on compressive strength of glass ﬁbre composites. Proceedings of the 18th Technical Conference, American Society for Composites, Gainesville, FL, Oct. 20–22. The Concrete Society (2000) Design Guidance for Strengthening Concrete Structures Using Fibre Composite Materials. TR55, 2nd edn. The Concrete Society, Camberley, UK. The Concrete Society (2003) Strengthening Concrete Structures Using Fibre Composite Materials: Acceptance, Inspection and Monitoring. TR57. The Concrete Society, Camberley, UK. Tomosawa F. and Nakatsuji T. (1997) Evaluation of ACM reinforcement durability by exposure tests. Proceedings of the 3rd International Symposium on Non-metallic (FRP) Reinforcement for Concrete Structures, Sappoto, 2, 139–146. Uomoto T. (2000) Durability of FRP as reinforcement for concrete structures. Proceedings of the 3rd International Conference on Advanced Composite Materials in Building and Structures, Ottawa, Ontario, Canada, 3–14. ICE Manual of Bridge Engineering # 2008 Institution of Civil Engineers Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. www.icemanuals.com 501 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved.