Structural Analysis of Historical Constructions - Modena, Lourenço & Roca (eds) © 2005 Taylor & Francis Group, London, ISBN 04 1536 379 9

Restoration and strengthening with fibre reinforced polymers: issues to consider

N.G. Shrive University ofCalgary, Calga/y, Canada

M.M. Reda Taha University of New Mexico, Albuquerque, USA

M.J. Masia The University of Newcastle, Newcastle, Australia

ABSTRACT: Fibre Reinforced Polymers (FRP's) have the advantages of light-weight, high strength and excellent durability which make them attractive for use in certain types of rehabilitation. However, there are disadvantages in terms of performance under heat and ultraviolet which need to be considered, and there is not much known about the long term bonding properties. Additionally, features peculiar to the use ofthese materiais need to be considered - a new mode of failure observed in reinforced beams, the modes of failure of FRP -wrapped columns and time dependent effects, including relaxation ofthe FRP due to flow ofthe bonding epoxy under shear.

TNTRODUCTION

Fibre Reinforced Polymers (FRP's) constitute a set of advanced composite materiais that have been consid­ered in recent years for various applications in the construction industry. While there have been appli­cations in new structures - some made entirely of FRP's (eg: Harvey 1993, Cooper 2004) - considerable effort has been placed on their use for rehabilitation and/or strengthening of reinforced concrete structures ([SIS Canada 200 I b, FIB 200 I). [n comparison, much less work has been performed with masonry (Lissel & Gayevoy 2003). With the latter, there has been some consideration of the potential use of FRP's in the restoration of historical structures.

There are some distinct advantages that FRP's would bring to a restoration project. Unidirectional FRP's in which the fibres (the ones of main interest are Glass (GRFP) and Carbon (CFRP» are aligned in the single direction during manufacture of the prod­uct, are very strong. (Some CFRP sheets, strips and rods have ultimate tensile strengths in the order of 3500 MPa.) The products are also very light and thus add little weight to the structure, potentially negat­ing the need to strengthen supporting members, and having negligible effect on the inertial response in

an earthquake. FRP's do not corrode like steel, and thus promise durability and longevity well beyond our expectations of unprotected steel. In order to achieve that durability, FRP's must currently be protected from ultraviolet light, which degrades the polymer matrix in which the fibres are embedded. New binders are being developed which are expected not to suffer from this problem.

A second area of development is in the stress­strain behaviour of these materiais. Currently they are essentially linear elastic to failure, with little ductility. Combi nations of fibres are now being investigated to produce materiais with non-linear behaviour and much greater strain capacity. The lack of ductility has led to reconsideration ofthe desired failure mode at ultimate. Whereas with steel, the aim is to ensure the steel yields before the concrete or masonry crushes; with FRP's it is the other way round. Sudden rupture of the FRP is to be avoided and warning offailure is to be obtained by producing relatively large deformations from sta­ble crushing ofthe compression material (ISIS Canada 200Ia).

In our view, there are additional issues which need to be considered when using FRP's in rehabilitation projects. We will touch briefly on the effects of heat and the durability of the bond between the FRP and

829

.l: C, c ~ ti O> c '2 .(õ

E ~ Õ ;f.

100

80

60

40

20

o -100

1\ . U Prestressing stee

J --CFRP - l eadline

\ L1 ~

\ I

100 300 500 700 Temperature °C

Figure I . Loss of strength of CFRP (Leadline) with tem­perature (Sayed-Ahmed & Shrive 1999).

the substrate (masonry or concrete); and then consider in more detail issues related to the fa ilure mode and creep in the rehabilitated structural element.

2 HEAT

The susceptibility of FRP's to fire is well known. However, there is only a small amount of information available. Fire retardant matrices are being developed so that an FRP will not burn of its own accord, sup­porting the fire. The response of these composites to heat needs to be studied, because heat alone can cause degradation in FRP properties (Sayed-Ahmed & Shrive 1999) . In Figure 1, we show that a CFRP pre­stressing tendon loses strength with temperature more quickly than steel. At 500°C, it took only two hours for the resin matrix to evaporate completely from around the fibres! However, recent developments ind icate much better performance can be achieved: the ten­sile strength of a carbon/geopolymer composite was reported to be 63% ofthe room temperature value after exposure to 800°C for one hour (Foden et aI. 1996) . Fire endurances equivalent to or better than conven­tionally reinforced concrete columns can be achieved with FRP-wrapped reinforced concrete columns, when appropriately designed and adequately protected (B isby et aI. 2004). No doubt calibrated numerical models can be used to extend to, and predict, other sit­uations and provide advice on the leveis of insulation required to meet necessary leveis of fire resistance.

3 DURABILlTY OF BOND

There has been some study ofthe durability ofthe bond between the FRP and the underlying substrate concrete or masonry (e.g. Green et aI. 2000). Initially, moisture alone could cause degradation: the effects offreezing and thawing have also been examined. Experiments are underway to study the effects of exposure under

different climate conditions of FRP's bonded to con­crete (Labossiere et aI. 2003). What is of interest is an accelerated test that simulates the long-term effects correctly. Debonding caused by environmental factors can not be afforded if FRP's are to Iive up to their potential as a long-term solution for strengthening and rehabi litation.

4 FAILURE MODES AND STRENGTH DESIGN

4. 1 General

Estimates ofhow much strength gain can be achieved by a particular application of FRP's are based on an assumed mo de of fai lure. The potentia l modes of fai lure are established from observations oftests. How­ever, it is not always clear what sequence of events is that leads to failure, and therefore what the limiting design criterion should be. For example, the fa ilure of masonry arches reinforced with FRP 's bonded to the intrados or extrados can be induced by one offour events - crushing of the masonry in compression, slid­ing between the units and the mortar due to shear, rupture ofthe FRP in tension, or debonding ofthe FRP if a mechanism develops in the arch (Briccoli Bati & Rovero 2000, Valluzzi et aI. 200 I). Which mechanism occurs depends on the material properties, the rei ative amounts ofmasonry and reinforcement, the geometry of the arch and the loading conditions. Models of lhe system can be used to assess which is the most likely mode of fai lure (Briccoli Bati & Rovero 200 1), and adjustments made to the leveis ofreinforcement iffur­ther refinement of the strength gain or type of failure (ductile/brittle) is des ired (Valluzzi et aI. 2001).

Similarly with masonry walls: considerable (thirty fold) increases in strength can be achieved, as well as large increases in deformations (Albert et aI. 200 1) through lhe application of externally bonded FRP. The modes of failure have been described and mod­eis have been developed to predict them (Kuzik et aI. 2003) . However, difficulties and inaccuracies can arise with strengthening iflhe actual mo de offai lure is not recognized. Two examples involving the addition of externally bonded FRPs are given below.

4.2 FRP strengthened RC beams

A substantial amount of work has been performed on strengthening reinforced concrete beams through the application of FRP strips to the soffi ts of the beams. This will provide increased flexural strength. Shear strength can be improved by the application of FRP sheets to the sides of beams. Design equations have been developed (eg. ISIS 2001 b, FIB 200 I) which can provide reasonable esti mates of the increases which can be expected, so long as the failure mode is as

830

Figure 2. Debonding from undernealh lhe load poinl back towards lhe support (Bakay 2003).

expected. One such mode in flexural strengthening is debonding ofthe laminate from its ends, typically near the support ofthe beams. (When a FRP strip is applied to the underside of the beam in situ , to strengthen the beam, it is usual to take the strip up to the support, but not lift the beam so the FRP can be taken in under the support.) Debonding occurs when the shear transfer to the strip is greater than can be carried by the bond between the strip and the concrete. Anchorage of the strip by additional means is required.

However, mid-span shear debonding has recently come to light as a new mode of failure (Riad et a!. 1998, Swamy et a!. 1999, Sayed-Ahmed et a!. 2004), Figure 2. In this case, debonding of the FRP strip begins under a load point near mid-span, at a load lower than expected and predicted for flexural failure (11 % gain in strength vs. 25% predicted (Sayed-Ahmed et a!. 2004)). The crack is initiated in the high moment plus shear zone at the edge of the high constant moment zone (in four-point bending). The debonding pro­gresses from under the load-point back toward the support - the reverse of the direction of the other debonding mode. This new mode of failure appears to be a result of insufficient strength in shear in the concrete between the flexural steel reinforcement and the FRP strip (Bakay 2003). As the beam is loaded ini­tially, strain compatibility occurs between the concrete in compression and the steel and FRP in tension. The loads carried in each material can be determined. Once the steel yields, any further increase in load causes an increase in the stress in the FRP only. Thus the con­crete between the steel and the FRP has to carry ali the additional shear from that which existed when the steel yields. The concrete beneath the steel reinforcement cracks in diagonal tension and the section ofthis con­crete in the centre of the beams displaces downward reI ative to the concrete on the other side of the crack. With the FRP in tension, there is thus a component of force pulling it off the concrete (Figure 3), resulting in debonding. An analytic model of this sequence of

831

, ,

• I

Figure 3. The debonding force develops from lhe angle of lhe FRP slrip.

events was able to predict the strengths observed for different beams with different concretes with accept­able agreement (Bakay 2003). The new mode of fail ure thus needed both to be recognized and understood before reasonably accurate predictions could be made ofits occurrence . The same sequence can also develop from a higher starting point in the beam section if there is inadequate shear reinforcement above the flexural steel.

4.3 FRP wrapped columns

Wrapping of columns with FRP's will only be a possi­ble solution in historical structures for cases where the column is not publicly visible, or where the column is parged, and fresh parging on 10p ofthe wrapping is acceptable. Estimates of the increase in strength that can be achieved are based either on the FRP wrap reaching its ultimate strain (ISIS Canada 200 I a) , or an analysis of published results and curve fitting on the conservative side of those results (Spoelstra & Monti 1999). For the FRP to confine a column actively, the FRP needs to be app lied prestressed. Otherwise the FRP relies on the column dilating near failure in order to be activated and resist further expansion of the column.

When a uniaxial fibre sheet is wrapped around a column, the direction of the fibres is circumferential, thus providing the maximum strength and stiffness to resist dilation of lhe column. When a wrapped, reinforced concrete column fails, a crack develops parallel to the direction of compression, splitting the fibres. The ultimate strain ofthe FRP has been reached across this split, but only locally (Pessiki et a!. 200 I , Scholefield 2003): the rest of the FRP is at strains well below ultimate. Failure occurs when the wrap rips circumferentially with cracks running between the fibres around the circumference of the column at both the top and the bottom of the longitudinal

Figure 4. Failure Unidirectional CFRP Jacket (Scholefield, 2003).

crack (Figure 4) (Scbolefield 2003). Basing column strengtb on a global acbievement of ultimate strain in tbe wrap, wben in fact it is a local condition, leads to an observed inconsistency in strengtb prediction (Shrive et aI. 2003). Recent evidence (Scbolefield 2003) suggests that lhe spalling of the concrete is driven by buckling oftbe reinforcement, complicating the issue furtber. Such a failure can not occur in plain masonry columns and indeed a different behaviour is observed. In this case, the mortar joints crusb and tbe units crack (Jongitudinally) into multiple pieces. Tbe column becomes rubble, contained witbin tbe FRP surround. Large strains can be accommodated witbout rupture of the bag (Masia & Shrive 2003), large stra ins can also be achieved witb FRP wrapped RC columns if tbe circumferential cracks are pre­vented from rapid propagation. This can be acbieved by using multi-directional sprayed FRP, or potenlially by applying FRP sbeets sucb tbat tbe fibres run cir­cumferentially in tbe one layer and longitudinally in another (Scbolefield 2003). The latter needs to be investigated.

Tbe majority of tests to date bave been performed on concentrically loaded colurnns. Tbe effect of eccen­tricity migbt be expected to be quite drastic, witb more rapid reduction in axial capacity witb increasing moment tban occurs witb a normal reinforced column. Tbe few tests tbat bave been performed (Cbaalal & Sbahawy 2000, Parvin & Wang 200 I, Li & Hada 2003) bave sbown failure develops through circumfer­ential tensile splitting of the FRP on the longitudinal tensi le face, and bursting of the concrete through tbe compression face . Bidirectional fabrics improved performance. Eccentricity in loading did reduce tbe confining effect on the columns.

There is thus a need to develop a method for predict­ing tbe strength increase tbat can be acbieved based on tbe actual mode of failure. The strain capacity is clearly influenced significantly by tbe mode tbree frac­ture toughness ofthe wrap combination used. Tbougbt

needs to be given as to tbe strengtbening tbat might be obtained witb eccentrically loaded columns, and bow tbat can be modeled and predicted.

5 CREEP

Tbe effects of creep in structures are not widely under­stood. Consider for example a concentrically loaded reinforced concrete column. Wben load is applied, we can apply tbe conditions of compatibility and equi­Iibrium to determine tbe stresses in tbe steel and tbe concrete. In time, tbe concrete creeps - wants to increase its compressive strain. Tbis means tbat to maintain compatibility, tbe strain in the steel must also increase. In tum, this means tbat as long as tbe steel is in tbe elastic part of its idealized stress- strain behaviour, tbe stress in tbe steel will increase. To maintain equilibrium, the stress in tbe concrete must tberefore decrease over time. Tbe effect of creep in tbis case therefore, is tbat tbe deformation increases and tbe stresses redistribute (steel stress increases, con­crete stress decreases) over time. Tbese are tbe two usual consequences of creep in any structure.

If tbe reinforcement is not elastic, but also creeps, tben the redistribution of stresses depends on tbe rela­tive speeds and amounts of creep in tbe two materiais. Stresses can botb rise and fali in time (Reda Taba & Shrive 2003). The redistribution of stress due to creep has been recognized as tbe likely cause of tbe col­lapse of the Civic Tower of Pavia (Shrive & Huizer 1991 , Binda et aI. 1992), and can cause cracking to develop in structures over time. lt is tberefore impor­tant to understand what may happen in a structure tbat is rehabilitated with FRPs.

WhileAramid and Glass FRP's creep, Carbon FRP's do not creep at normal service stresses. However, tbere is another source of potential time-dependent bebaviour in tbe system - creep oftbe epoxy bonding agent between the FRP and tbe underlying structure. We bave begun to examine the possible effect of tbis by loading two beams in flexure (Masia et aI. 2004). The beams bave tbe same internai steel reinforcement and were cast from the same batch of concrete. One oftbe beams bad two CRFP strips glued to the tension side. The beams were loaded one after tbe other on tbe same day in four-point bending. As expected, tbe cen­traI deflection of the FRP reinforced beam was much lower than that of tbe beam without the strips. The central displacements of tbe beams bave been moni­tored over time and are shown in Figure 5. Botb beams creep as expected. In the beam without the FRP strips, the concrete creeps whicb increases tbe curvature and lhe deflection. Tbe increased curvature increases the stress in the steel. This in tum requires an increase in concrete force to maintain zero total axial force . However, tbe moment is unchanged so the lever arm must decrease. Tbus the concrete stress decreases, but

832

Ê 8,--------------------,

.s 7 +------------------,---1 I:

.g 6 +--------=".....,=----=--=--~---I ~ ~ 5 +---."..6~;.-.....~'-'-'.c...:..'--==_----j

[4 +-~F------~~------~ '" ~ 3 -hi'T---,_'=.-"--~~=-==:..:------- ----------- -1

E 2f7~~~~?=~~~====~~~~ $ ~ 81 Exp ---- 82 CE8-FIP g> 1 -li"--------I~ 82 Exp - - - . 81 ACI .3 O - - 81 CE8-FIP ...... 82 ACI

O = ~ ~ ~ 1=1= 1~1~ Time afie. loading (days)

Figure 5. Long term midspan deflection versus time after loading (1417 days - ali data shown) and predictions using different creep models (Masia et aI. 2004).

the area of concrete which is loaded increases as the neutral axis moves down in the beam and the lever arm is thus reduced. The effective modulus ofthe concrete is thus decreasing over time and the second moment of area of the transformed section is also a function oftime.

The same sequence of events happens in the beam with FRP strips. The presence of the strips makes the beam stiffer initially - sufficiently stiff so that the concrete did not crack in tension. As the curvature increases from the creep of the concrete the tensile strain increases and there comes a time when the con­crete will begin to crack in the tension zone. This was observed in the tests: a flexural crack in the concrete was noticed several months after loading. This will cause a change in the second moment of area in addi­tion to the shifting neutra I axis. Another feature which adds complexity to the problem is relaxation of the FRP stress due to flow of the epoxy binding agent gluing the FRP to the beam. Shear occurs across the interface in order for tension to develop in the FRP. Such shear occurs in any shear zone ofthe beam, where the moment is changing. We attempted to measure slip of the FRP from this mechanism with simple spring gauges touching the ends of the strips. Some move­ment (off loading the FRP) was measured in the first few weeks after loading (Figure 6).

The creep of the beam was modeled using the approaches in the CEB-FIP model code (1990) and the ACI Committee 209 recommendations (ACI 1992), obtaining a best fit for the first beam (no FRP). Creep for the second beam (with FRP strips) was predictecl, allowing for the increase in stiffness from the FRP but assuming no flow in the epoxy binding agent. The creep is underpredicted at longer ages (Figure 5). Whether this is due to the change in second moment of area (from uncracked to cracked section), to the relax­ation of load in the FRP due to flow in the epoxy, or to a combination thereof is under analysis. Initial results indicate the effect is due to the combination offactors.

Ê 0.12 ,------------------,

.s {l 0.1 +---------=.....,,::=--------{ I: Q)

.g- 0.08 +--~----tI'-------------l 1ii Q.

ff: 0.06 -thi;i~~II'--'--==-~==~-==~-I u ... . 9- 0.04 -If:lr------------------l Ui ~ 0.02 'f-----------j :; Qj ~ O~-.--.--._-._-r_-r_-r___l

o = ~ ~ ~ 1=1=1~1~

Time afie. loading (days)

Figure 6. Relative slip versus time at each end of one CFRP strip (beam 82) (1417 days) (Masia et aI. 2004).

Notwithslanding lhe cause, the measurements show the FRP does unload to some degree, certainly in the initial days after load application. For historical struc­tures, if FRP is applied pre-stressed to counteract an effect of dead-Ioad or a permanent live loacl, some loss of this action must be anticipated from bond relax­ation, unless the ends of the FRP strips are securely fastened by means other than epoxy bond. Note the beam here had additional shear capacity from the plain beam through the application of GFRP sheets in the shear sections of the span. The sheets were wrapped from one side, over the FRP strips and up the olher side of the beam. The relaxation occurred despite this additional anchorage. Hence a more rigid means must be used (eg. Bolts through the strips into the concrete) to avoid the relaxation.

6 CONCLUSION

The use of FRP's in restoration and rehabilitation of historic structures is an attractive option. Large gains in strength can be achieved for some situations with minimal additional weight. The need to "hide" the FRP is an obvious consideration in any scheme, but there are other features that need to be considered in the full assessment of a scheme: the effects of creep and relaxation of the FRP through bond flow; the need for toughness in the FRP arrangement to reduce crack propagation and rapid failure; the need to ensure that the mode of failure is understoocl, and that allmodes have been considered; and that suitable precautions are taken to obtain sufficient fire resistance and durability in the bond.

ACKNOWLEDGEMENT

The authors gratefully acknowledge the support they have had from their respective institutions in formu­lating these ideas, and the support ofISIS Canada, The

833

Natural Sciences and Engineering Research Council of Canada, and the Universities ofCalgary and Newcastle for the various experiments performed.

REFERENCES

ACI 1992 Manual of Concrete Practice, Part I, 1992. Prediction of creep, shrinkage and temperature effects in concrete structures, Standard ACI 209 R-92, American Concrete lnstitute .

Albert, M.L., Elwi, A.E. & Cheng, J.J.R. 200 I. Strengthening of Unreinforced Masonry Walls Us ing FRP's, JOllrnal of Compositesfor Constrllction, ASCE, 5( I), 76- 84.

Bakay, R.C.P. 2003. Midspan Shear Debonding of CFRP Laminated Reinforced Concrete Beams. MSc Thesis, University ofCalgary.

Binda, L., Gatti, G., Mangano, G., Poggi, C., Sacch i, & Landriani, G. 1992. The collapse of the civic tower of Pavia: A survey of the materia is and structure, Masonry International, 6(1), 11-20.

Bisby, L.A., Kodur, VK.R. & Green, M.F. 2004. Numeri­cal Parametric Studies on the Fire Endurance of FRP­Confined Concrete Columns, Canadian Journal ofCivil Engineering, 31, in press.

Briccoli Bati, S. & Rovero, L. 2000. Consolidation of Masonry Arches with Carbon-Fiber Reinforced Plastics, Proc. i2th Int 'I BricklBlock Masonry Conference, Madrid, Spain, 1571- 1582.

Briccoli Bati, S. & Rovero, L. 200 I. Experimental Va lidation of a Proposed Numerical Model for lhe FRP Consolidation ofMasonry Arches. Historical Constrllction - Proc. 3rd Int 'I Seminal', Guimaraes, Portugal, 1057-1066.

CEB-FIP 1990. CEB-FIP Model Co de 1990 - Design Code Comité Euro-Interantional du Beton, Thomas Te lford, London, UK.

Chaalla l, O. & Shahawy, M. 2000. Performance of Fibre-Reinforced Polymer-Wrapped Reinforced Concrete Columns Under Combined Axial and Flexural Loading, ACI Structural Journal, 97(4) 659- 668.

Cooper, D.1. 2004. Halgavor Bridge - A Case Study - CD Proc. 4th Int 'I Conf on Advanced Composite Materials in bridges and Structures, Calgary, Canada, 8pp.

FIB, CEB-FIP 200 I. Externally Bonded FRP Reinforcement For RC Structures, Technical Report, Bulletin 14.

Foden, A. , Lyon, R., Balaguru, P. & Davidovitz, J. 1996. High Temperature lnorganic Resin For Use in Fiber Rein­forced Composiles. Proc. 1st Int'! Conf on Composites in Inji-astructllre. Dept. Civ. Eng. Univ. Arizona, USA, 166-177.

Green, M.F., Bisby, L.A. , Beaudoin, Y. & Labossiere, P.2000. Effect of Freeze - Thaw Cycles on the Bond Durability Between Fibre Reinforced Polymer Plate Reinforcement and Concrete, Canadian JOllrnal of Civil Engineering, 27(5), 949- 959.

Harvey, WJ. 1993. A Reinforced Plastic Footbridge, Alber­feldy, UK. Strllctural Engineering International, 3(4), 229- 232.

ISIS Canada 2000a. Reinforcing Concrete Structures with Fibre Reinforced Polymers. Design Manual No. 3.

ISIS Canada 200 I b. Strengthening Reinforced Concrete Structures with Externally-Bonded Fibre Reinforced Polymers. Design Manual No. 4.

Kuzik, M.D. , Elwi, A.E. & Cheng, J.J.R. 2003. Cyclic Flex­ure Tests of Masonry Walls Reinforced with Glass Fiber Reinforced Polymer Sheets, Journal of Composites for Construction, ASCE, 7( 1), 20- 30.

Labossiere, P., Neale, K. W & Nishizaki, 1. 2003. Effects of Different Long-Term Climatic Conditions on FRP Dura­bi lity, Proc. 6th int 'l. Symp. On Fibre Reinjàrced Polymer (FRP) Reinforcementfor Concrete Structures, Singapore, Vo l. 2.

Li, 1. & Hadi, M.N.S. 2003. Behaviour of Externally Con­fined High Strength Concrete Columns Under Eccentric Loading. Composite Structures, 62(2),145- 153 .

Lisse l, S.L. & Gayevoy, A. 2003. The Use of FRP's in Masonry: A State ofthe Art Review. Proc. Int 'I. Conf on lhe Performance ofConstruction Materiais, Cairo, Egypt, 1243-1252.

Lourenco, P.B. & Pocas Martins, 1.P. 200 I. Strengthening of the Architectural Heritage with Composite Materiais. Proc. oflnternational Conference ofComposites in Con ­struction, Figueiras et al. (Eds), Porto, Portugal, Swets and Zeitlinger, 57 1- 576.

Masia , M.J. & Shrive, N.G. 2003 . CFRP Wrapping for the Rehabilitation of Masonry Columns, Canadian Journal ofCivil Engineering, 30(4), 734- 744.

Masia, M.J., Shrive, N.G. & Shrive, P.L. 2004. Creep Behaviour of RC Beams Strengthened With Externally Bonded FRP Strips, Proc. 18th Australasian Conference on the Mechanics of Structures and Materials, Perth, (submitted).

Parvin, A. & Wang, A. 2001. Behaviour of FRP Jacketed Concrete Columns Under Eccentric Loading, Journal of Composites for Construction, ASCE, 5(3), 146- 152.

Pessiki , S., Harries, K., Kestner, 1. , Sause, R. & Ricles, J. 200 I. Axia l Behaviour of Reinforced Concrete Columns Confined with FRP Jackets,ASCE Journal ofComposites for Constrllction, 1(3),237- 245.

Reda Taha, M.M. & Shrive, N.G. 2003. Numerical Modelling of Non-Linear Creep Effects on lhe Stress Distribu­tion in Composite Structures. Proc. 9th Arab Structural Engineering Conf, Emerging Technologies in Structural Engineering, Abu Dhabi, UAE, 47- 54.

Riad, A., Sayed-Ahmed, E.Y., Tadros, G., Loov, R.E. & Shrive, N.G. 1998. Design Recommendations for HC­Type Bridge Girders Retrofitted With CFRP Strips, CD Proc. 5th Int 'l. Conf on Short and Medium Span Bridge Engineering, Calgary. Canada,CSCE, 1175- 1188.

Sayed-Ahmed, E.Y. & Shrive, N.G. 1999. Smart FRP Pre­stressing Tendons: Properties and Prospects. Proc. Sec­ond Middle East Symposium on Structural Compos­ites for Infrastructure Application. Hurghada, Egypt, 80- 93.

Sayed-Ahmed, E.Y., Riad, A.H. & Shrive, N.G. 2004. Flexu­ral Strengthening of Precast Reinforced Concrete Bridge Girders Using Bonded Carbon Fibre Reinforced Polymer Strips or Externai Post-Tensioning. Canadian Journal of Civil Engineering, 31, 14pp, in press.

Scholefield, B.WJ. 2003. Rehabi litation of Reinforced Con­crete Columns by Various Methods, MSc Thesis, Univer­sity ofCalgO/y.

Shrive, P.L., Azarnejad, A., Tadros, G., McWhinnie, C. & Shrive, N.G. 2003. Strengthening of Concrete Columns with CFRP wrap. Canadian Journal ofCivil Engineering, 30(3), 543- 554.

834

Shrive, N.G. & Huizer, A. 1991. TMS Open Forum, The Masomy Society JOllrnal, 9(2), 4.

Spoelstra, M.R. & Monti , G. 1999. FRP Confined Concrete Model, ASCE JOllrnal of Composites for Constrllction, 3(3), 143- 150.

Swamy, R.N. & Mukhopadhyaya, P. 1999. Debonding of Carbon-Fibre-Reinforced Polymerplate Concrete Beams.

835

Proc. /nstituteofCivil Engineers, Strllctllres and BlIilding, 134(4), 301 - 317.

Valluzzi , M.R. , Valdemarca, M. & Modena, C. 2001. Behaviour of Brick Masonry Vaults Strengthened by FRP Laminates. JOllrnal ofCompositesfor Constrllction, ASCE, 5,163- 169.