guide_bs

69
1 Design Guide Line for S&P FRP Systems Software: www.reinforcement.ch www.frp.at Note: The design models and the "FRP Lamella" software are based on the mate- rial parameters of the S&P reinforcing fibres and of the S&P adhesive systems. If other components are used, the required FRP cross section and the anchor check provided by the software will no longer be valid. Under these circumstances the system supplier waives all liability. The current updated version of the software can be found on the Internet at www.reinforcement.ch or www.frp.at .

description

FRP

Transcript of guide_bs

1

Design Guide Line for

S&P FRP Systems

Software: www.reinforcement.ch www.frp.at Note: The design models and the "FRP Lamella" software are based on the mate-rial parameters of the S&P reinforcing fibres and of the S&P adhesive systems. If other components are used, the required FRP cross section and the anchor check provided by the software will no longer be valid. Under these circumstances the system supplier waives all liability. The current updated version of the software can be found on the Internet at www.reinforcement.ch or www.frp.at.

2

Page 1. Introduction 4 2. FRP reinforcing systems 4 3. Types of fibres for S&P FRP systems 6 4. S&P FRP sheets for site lamination 7 4.1 S&P A-Sheet 120 8

4.2 S&P C-Sheet 240 9 4.3 S&P C-Sheet 640 10 4.4 S&P G-Sheet 50/50 11 4.5 S&P G-Sheet 90/10 Type A 12 4.6 S&P G-Sheet 90/10 Type B 13

5. Cold curing epoxy resin systems 14 6. S&P FRP systems using prefabricated laminates 15 7. Prestressed FRP systems 16 8. Substrate requirements for FRP applications 17 9. Bonding and end anchoring of FRP systems 18 9.1 Confinement of columns 18

9.2 Design for flexural strengthening using S&P Laminates CFK 18 9.3 Anchoring at the ends of the laminates 20 9.4 Prestressed FRP systems 21

9.4.1 Prestressed S&P Laminates CFK 21 9.4.2 Prestressed S&P A-Straps 22

9.5 Slot application of S&P Laminates CFK 23 10. Selection of the S&P FRP system 24 10.1 Axial load confinement of columns 26

10.1.1 General 26 10.1.2 Evaluation of the axial compressive load – axial deformation response 26 10.1.3 Determination of the compressive strength of the confined concrete 29 10.1.4 Evaluation of the lateral confining pressure 30 10.1.5 Design example 31

10.2 Prestressing of axially loaded columns 34 10.3 Combined moment and axial load on columns 35 10.4 Replacement of corroded reinforcement using FRP 35 10.5 Seismic upgrading of columns 37 10.6 Flexural strengthening with non-prestressed FRP systems 37

10.6.1 Surface bonded S&P Laminates CFK 38 10.6.2 S&P C-Sheet 240 38 10.6.3 Slot-applied S&P Laminates CFK 39 10.7 External shear reinforcement 41

10.8 Flexural strengthening with prestressed S&P Laminates CFK 43 10.8.1 Testing arrangement 44 10.8.2 Test results 45

3

10.8.3 Conclusion 45 10.9 Special applications using FRP systems 46 10.9.1 Repairs to historic buildings 46 10.9.2 Applications in timber construction 47 10.9.3 Fire protection measures for FRP strengthened elements 48

11. Seismic retrofitting of reinforced concrete columns and frames using FRP 49

11.1 Reinforced concrete columns 49 11.1.1 Strengthening methods for reinforced concrete columns 49 11.1.2 Reinforced concrete columns retrofitted with FRP jackets 49 11.1.3 Confinement to improve flexural ductility 50 11.1.4 Confinement to improve the lap splices of the internal reinforcement 52 11.1.5 Confinement for external shear strengthening 53 11.1.6 Flexural strength of the column 54 11.1.7 Comparison of strengthening methods 54 11.2 Reinforced concrete frames 55

12. Explosion and impact protection using S&P FRP systems 57

12.1 Explosion protection 57 12.2 Impact protection 58

13. Seismic retrofitting of masonry using S&P FRP systems 60

13.1 Masonry strengthening using the S&P G-Sheet AR 50/50 60 13.2 Masonry strengthening using S&P Laminates CFK 61 13.3 Load bearing wall systems 61

14. Texts for specifications 62

14.1 Flexural strengthening with S&P Laminates CFK 62 14.2 Reinforcement of load bearing columns with S&P C-Sheet 240 63 14.3 External shear reinforcement with S&P C-Sheet 640 64 14.4 Reinforcement with S&P G-Sheet AR 50/50 (poor substrate / historic buildings) 65 14.5 Seismic retrofitting of columns 65 14.6 Flexural strengthening with prestressed S&P Laminates CFK 66

15. S&P FRP product overview 67 16. Literature / Product names 68

4

1. Introduction The structural design and thus the production of structural elements made of reinforced concrete is based on forces and loads current in codes of the time. However, during the service life of a structure, various circumstances may require that the service loads are changed due to: • Modification of the structure • Aging of the construction materials • Deterioration of the concrete caused by reinforcement corrosion • Earthquake design requirements; fire design changes • Upgrading of building codes relating to load bearing capacity or service loads etc. The consideration of the actual loads resisted by a structure is a necessary prerequisite for the development of a comprehensive rehabilitation concept. Basically, the methods available for the structural strengthening of structural elements made of reinforced con-crete are as follows: • Application of cast or sprayed concrete with additional reinforcement • Placing of additional reinforcement in slots cut in the concrete • External post-tensioning • Installation of supports or additional bearing members • Steel plate bonding to increase shear or flexural capacity. An alternative to these traditional methods of strengthening is the use of Fibre Rein-forced Polymer (FRP) composites. 2. FRP reinforcing systems Reinforcing steel and fibre composites exhibit different material behaviours. While rein-forcing steel shows an ideal elastic-plastic behaviour, all FRP systems are linear-elastic materials. This circumstance must be taken into account in design and dimensioning. The basic fibres of FRP systems are imbedded in a polymer matrix and their arrange-ment can be either uni-directional or bi-directional. FRP composites are used for the ret-rofit of existing structural components. Stretched sheets Uni-directional arrangement: The fibres are bonded to a tight mesh and supplied as a sheet. The parallel fibres in the sheet are stretched and thus provide a high modulus of elastic-ity. They are especially suited for increasing the structural capacity of an element.

5

Woven sheets Bi-directional arrangement: The arrangement of the fibres is bi-directional. The sheets are produced by weaving. In the woven sheets the fibres become slightly wave like in form. As a result, these products are less suited for increasing the structural capacity of an element. Bi-directional sheets are ideally suited for increasing the ductility of a building component. Under loading, the fibres must be stressed before higher forces can be absorbed. Uni-directional and bi-directional FRP sheets Cold or thermally curing epoxy resins matrices are used to ensure load transfer from the sheets to the substrate. Cold curing epoxy resin matrix: Up to a weight of 400 gm/m2 both uni-directional and bi-directional sheets are applied as a dry lay up. This means: The cold curing epoxy resin is rolled on to the structural element. The dry sheet is then applied into the matrix. Stretched and woven sheets with a weight of 400-800 gm/m2 are applied as a wet lay up. This means: The sheets are impregnated with the epoxy matrix and applied wet to the building com-ponent. Thermally curing epoxy resin matrix: The uni-directional and bi-directional sheets are impregnated with the thermally curing epoxy adhesive and delivered at controlled low temperatures. Thermal curing is done by applying heat to the epoxy resin on the element. This type of reinforcement is called Prepreg and is typically used in the aviation and sports industry; for the retrofit of struc-tures Prepreg is normally not used. Prefabricated FRP (laminates) Prefabricated FRP is delivered to the job site as a composite (laminate). Impregnation with the matrix and thermal curing is done by the supplier under controlled factory condi-tions. S&P offers uni- and bi-directional laminates to the individual requirements of the application. The most well known laminate used for structural strengthening is the S&P Laminate CFK.

6

3. Types of fibres for S&P FRP systems

Type of fibre Modulus of elasticity GPa

Tensile strength MPa

C (Carbon) 240 - 640 2’500 – 4’000 A (Aramid) 120 3’000 – 4’000 G (Glass) 65 - 70 1’700 – 3’000 Polyester 12 - 15 2’000 – 3'000

Figure 1: Stress-strain-diagram

(Steel 210 550)

S&P manufactures custom-made sheets of either a single fibre type or of fibre combina-tions (hybrids). The advantages and disadvantages of the various fibres are as follows: E-glass: Uncoated E-glass corrodes in alkaline environments. Since the coating

may be subjected to damage near its edges during the application, E-glass should only be applied in non-alkaline areas (reinforcement of asphalt) or where it is completely embedded in the epoxy resin matrix. E-glass sheets should not be used for strengthening of concrete in combination with a wa-ter vapour-permeable matrix like S&P Resicem.

AR-glass: Alkali-resistant glass is suited to confinement reinforcement for concrete elements (increase in ductility) in combination with S&P Resicem, a water vapour-permeable matrix.

An intensive research program has been carried out with the S&P AR-glass. During 28 days the glass was stored in the substances listed below and then examined: Saturated solutions of calcium hydroxide: 10% potassium hydroxide 10% sodium hydroxide Acids: Chlorides: 5% acetic acid 10% calcium chloride 10% hydrochloric acid 10% sodium chloride 10% nitric acid 10% potassium chloride Nitrates: Sulphates: 10% calcium nitrate 10% calcium sulphate 10% potassium nitrate 10% sodium sulphate 10% sodium nitrate 10% potassium sulphate10% ammonium nitrate

Figure 2: View of S&P AR-glass in the scanning electron microscope, after storage

The S&P AR-glass was resistant against all acids, salts and alkalis examined in the test. After a storage of 28 days none of the samples showed any weight reduction. No sur-face modification and thus no attack was observed in the scanning electron microscope.

7

The S&P AR-glass is highly alkali-resistant (resistance against 100% caustic soda). Aramid: Aramid is a very tough material. Aramid sheets are used, as an example,

for the manufacture of bullet-proof vests. For structural elements this ex-treme toughness provides benefits for special applications such as the strengthening of rectangular columns. Ideally, aramid fibres used for this application are prestressed prior to the application. Aramid is especially suited as impact protection for the retrofit of columns.

Carbon fibres: Carbon fibres provide special benefits when used to increase the flex-

ural capacity:

• High modulus of elasticity (depending on fibre type) • Minimum coefficient of thermal expansion (approx. 50 times lower-

than steel) • Excellent fatigue properties • Excellent resistance to all types of chemical attack • Will not corrode • Freeze/thaw and de-icing salt resistance. 4. S&P FRP sheets for site lamination The S&P sheets (uni-directional/bi-directional) can be applied as dry and wet lay ups and also as preimpregnated Prepegs. The function of the matrix is to enable the load transfer between the fibres and from the fibres to the substrate. Dimensioning of the re-inforcement is based on the theoretical fibre cross section and the theoretical fibre pa-rameters only. The theoretical sheet thickness is determined as follows:

Theoretical sheet thickness = fibre weight in the reinforcement direction

density of the fibre Due to the application process, the site lamination does not always produce an optimum fibre arrangement. There is also a risk of damage to the fibres while they are being rolled on to the surface. It is therefore recommended that the E-modulus used for the fi-bres be reduced by an environmental reduction [y]. The environmental reduction factor is based also on the durability of each fibre type.

Recommended reduction factor [y]: uni-directional (stretched) S&P C-Sheets y = 1.1-1.2 uni-directional (stretched) S&P A-Sheets y = 1.2-1.4 bi-directional (woven) S&P G-Sheets y = 1.5-1.8

The theoretical parameters of the S&P sheets for site lamination are as follows.

8

4.1 S&P A-Sheet 120 Sheet of aramid fibre for dry or wet lay up

Technical data of fibre (uni-directional)

290 gm/m2 420 gm/m2 650 gm/m2

Modulus of elasticity (GPa) 120 120 120

Tensile strength (MPa) 2900 2900 2900

Weight of A fibre (gm/m2) (main direction)

290 420 650

Total weight of sheet (gm/m2) 320 450 680

Density (gm/cm3) 1.45 1.45 1.45

ε ultimate % 2.5 2.5 2.5

Thickness for statical design weight/density (mm)

0.20 0.29 0.45

Theor. section for statical design 1000 mm width (mm2)

200

290

450

Reduction factor for statical design (manual lamination / UD-product)

1.3 (recommended by

S&P)

1.3 (recommended by

S&P)

1.3 (recommended by

S&P)

Tensile force of 1000 mm width for design (kN)

200 x 2900 = 446.2 1.3

290 x 2900 = 646.9 1.3

450 x 2900 = 1003.8 1.3

Application: •

Impact protection • Explosion protection

9

4.2 S&P C-Sheet 240 Sheet of carbon fibre for dry lay up

Technical data of fibre (uni-directional)

200 gm/m2 (400 gm/m2) 300 gm/m2

Modulus of elasticity (GPa) 240 240

Tensile strength (MPa) 3800 3800

Weight of C fibre (gm/m2) (main direction)

200 (400) 300

Total weight of sheet (gm/m2) 230 (430) 330

Density (gm/cm3) 1.7 1.7

ε ultimate % 1.55 1.55

Thickness for statical design weight/density (mm)

0.117 (0.234) 0.176

Theor. section for statical design 1000 mm width (mm2)

117 (234)

176

Reduction factor for statical design (manual lamination / UD-product)

1.2 (recommended by S&P)

1.2 (recommended by S&P)

Tensile force of 1000 mm width ultimate (kN)

117 x 3800 = 370.5 1.2 (741)

176 x 3800 = 557.3 1.2

Tensile force of 1000 mm width at 0.6 % εεεε for design (kN)

140 (280) 211

Application: • Axial load enhancement of circular and square cross-sections containing a large amount of internal stirrup reinforcement

Flexural strengthening at low reinforcement levels or with substrates with a low inherent tensile strength

Note: Different weight of S&P C-Sheet 240 upon request.

10

4.3 S&P C-Sheet 640 Sheet of carbon fibre for dry lay up

Technical data of fibre (uni-directional)

400 gm/m2

Modulus of elasticity (GPa) 640

Tensile strength (MPa) 2650

Weight of C fibre (gm/m2) (main direction)

400

Total weight of sheet (gm/m2) 430

Density (gm/cm3) 2.1

ε ultimate % 0.4

Thickness for statical design weight/density (mm)

0.19

Theor. section for statical design 1000 mm width (mm2)

190

Reduction factor for statical design (manual lamination / UD-product)

1.2 (recommended by S&P)

Tensile force of 1000 mm width ultimate (kN)

190 x 2650 = 419.6 1.2

Tensile force of 1000 mm width at 0.2 % εεεε for design (kN)

200

Application: • External shear strengthening • End anchoring of the S&P laminate CFK

11

4.4 S&P G-Sheet E 50/50, S&P G-Sheet AR 50/50 Sheet of E- or AR-glass fibre for dry lay up

Technical data of fibre (main- and cross direction)

E-glass AR-glass

Modulus of elasticity (GPa) 73 65

Tensile strength (MPa) (virgin filament)

3400 3000

Impregnated strand in a composite (MPa)

2400 Use for the static design

1700 Use for the static design

Sheet weight (gm/m2) (total 350 gm/m2)

175 in both directions

175 in both directions

Density (gm/cm3) 2.6 2.68

ε ultimate % 4.5 4.3

Thickness for statical design weight/density (mm)

0.067 0.065

Cross section for statical design 1000 mm width (mm2)

67 (fibre area only/each direction)

65 (fibre area only/each direction)

Reduction factor for statical de-sign (manual lamination / woven product)

1.6 (recommended by S&P)

1.6 (recommended by S&P)

Tensile force of 1000 mm width for design (kN)

67 x 2400 = 101.00 1.6 each direction

65 x 1700 = 69.00 1.6 each direction

Application: • Explosion protection •

Reinforcing of masonry or historic buildings • Seismic retrofitting

12

4.5 S&P G-Sheet E 90/10 A, S&P G-Sheet AR 90/10 A Sheet of E- or AR-glass fibre for dry lay up

Technical data of fibre (main direction)

E-glass AR-glass

Modulus of elasticity (GPa) 73 65

Tensile strength (MPa)(virgin filament)

3400 3000

Impregnated strand in a composite (MPa)

2400 Use for the static design

1700 Use for the static design

Sheet weight (gm/m2)(total 440 gm/m2)

400 in main direction

400 in main direction

Density (gm/cm3) 2.6 2.68

ε ultimate % 4.5 4.3

Thickness for statical design weight/density (mm)

0.154 0.149

Theor. section for statical design 1000 mm width (mm2)

154 (fibre area only main direction)

149 (fibre area only main direction)

Reduction factor for statical design (manual lamination / woven product)

1.6 (recommended by S&P)

1.6 (recommended by S&P)

Tensile force of 1000 mm width for design (kN)

154 x 2400 = 231.00 1.6 main direction

149 x 1700 = 158.00 1.6 main direction

Cross direction 10 % of the equal fibre is used in the weft (cross section)

Application: • Seismic retrofitting of columns using dry lay up.

13

4.6 S&P G-Sheet E 90/10 B, S&P G-Sheet AR 90/10 B Sheet of E- or AR-glass fibre for wet lay up

Technical data of fibre (main direction)

E-glass AR-glass

Modulus of elasticity (GPa) 73 65

Tensile strength (MPa) (virgin filament)

3400 3000

Impregnated strand in a composite (MPa)

2400 Use for the static design

1700 Use for the static design

Sheet weight (gm/m2) (total 880 gm/m2)

800 in main direction

800 in main direction

Density (gm/cm3) 2.6 2.68

ε ultimate % 4.5 4.3

Thickness for statical design weight/density (mm)

0.308 0.299

Theor. section for statical de-sign 1000 mm width (mm2)

308 (fibre area only main direction)

299 (fibre area only main direction)

Reduction factor for statical design (manual lamination / woven product)

1.6 (recommended by S&P)

1.6 (recommended by S&P)

Tensile force of 1000 mm width for design (kN)

308 x 2400 = 462.00 1.6 main direction

299 x 1700 = 318.00 1.6 main direction

Cross direction: 10 % of the equal fibre is used in the weft (cross section)

Application: • Seismic strengthening of columns using wet lay up.

14

5. Cold curing epoxy resin systems

Figure 3: A structure must be permeable to water vapour

from the inside to the outside. VITRUV (Roman architect and engineer, approx. 50 B.C.)

ACI FRP Guideline 440 (Section 8.3.3 Durability) The effective US Guideline 440 for external FRP strengthening prescribes: Any FRP system that com-pletely encases or covers a concrete section should be in-vestigated for vapour pres-sures, and moisture vapour transmission.

When total surface wrapping of concrete is intended, aspects of building physics must be considered. 30-50% of the surface of the element should remain water vapour-permeable. A total surface coverage with an epoxy matrix is therefore not suitable. S&P Resicem is a newly developed cementitious epoxy matrix. The combined effect of the two binders that have completely different chemical bases, is that the cement parti-cles, due to water vapour pressure, penetrate into the microstructure of the epoxy resin. Thus, the matrix system, which is vapour-proof at the time of its application, becomes vapour-permeable as the water vapour exposure increases. The cement contained in the matrix provides an additional alkali deposit which protects the internal reinforcement against corrosion. The water vapour diffusion coefficient of a FRP confinement (thick-ness 1 mm) with S&P Resicem will eventually level out at approx. 3’000-5’000. Applica-tion is possible to substrates with a moisture content of up to 12%. Determination of the water vapour-permeability of a coating Sd = µµµµH2O x layer thickness (m) < 4 m Sd: Water vapour diffusion resistance of the layer µµµµH2O: Water vapour diffusion coefficient of the coating S&P Resicem µH2O ≅ 3’000 – 5’000 (FRP thickness 1 mm) S&P Resin Epoxy 50/55 (epoxy resin) µH2O = 1’000’000 (FRP thickness 1 mm) It follows that: Sd: of a two-layer FRP coating using S&P Resicem, with a total thickness of 0.8 mm (matrix + fibres), Sd = 4’000 x 0.008 (m) = 3.2 m < 4 m With the use of S&P Resicem the water vapour-permeability of the FRP system is guar-anteed.

15

6. S&P FRP systems using prefabricated laminates Prefabricated laminates are produced by the extrusion method. In a continuous process, the carbon fibres are soaked in epoxy resin and cured through heating. For technical reasons, the extrusion method limits a maximum fibre content to approx. 70%. The elas-tic properties of a uni-directional composite can be calculated from the performance of the fibres and of the matrix. Since the modulus of elasticity and the tensile strength of the matrix can be neglected for the calculation of the laminate properties, the values are approx. 70% of the values of the carbon fibre. The S&P Laminates CFK are manufac-tured by a new method. For the production of S&P hybrid laminates various types of carbon fibres with different moduli of elasticity and different tensile strengths are used. Normally, the modulus of elasticity of a hybrid laminate does not have a linear progres-sion. Carbon fibres with a high modulus of elasticity and a low ultimate elongation will break sooner than fibres with a low modulus of elasticity. In the hybrid, however, the C fibres with a higher elongation are prestressed during the production procedure. This produces a linear progressive modulus of elasticity in the laminate. Due to this hybrid technology, inexpensive C fibres with a low modulus of elasticity can be used. While the design for manual on-site lamination is based on the theoretical fibre cross section and the parameters of the fibres only, the design for application of prefabricated S&P Laminates CFK is based on the cross section and the parameters of the compos-ite.

Figure 4: Installation of CFK-

Laminates

S&P Laminates CFK

The parameters of the laminates are checked by S&P according to the ISO 9001 quality assurance concept. Thus, conversely to manually applied FRP systems, the E-moduli of the laminates do not need to be re-duced by an additional reduction factor (y = 1).

Properties S&P Laminate 150/2000 S&P Laminate 200/2000

Modulus of elasticity for design 164’000 MPa 205’000 Mpa

Ultimate tensile strength 2’700 – 3’000 MPa 2’400 – 2’600 MPa Special laminates with a modulus of elasticity of 300’000 MPa can be custom-made to the specific requirements of a project. However, the application is often not economical because of their low effective tensile strength.

16

7. Prestressed FRP systems FRP systems can basically be applied in the non-prestressed or prestressed condition. Preliminary tests show that softer materials (glass or aramid fibres) tend to be more suitable for prestressing than high modulus carbon fibres. A high modulus of elasticity of the fibre on the one hand means a lower elongation for a defined prestress level and, on the other, a tendency of premature failure at reduced eccentricities during the prestress-ing procedure. In order to transmit the tensile forces, prestressing requires the FRP sys-tems to be fixed (locked) at both ends. This is much more difficult with the brittle C and G fibres than with the tough aramide fibres. Two systems are available for prestressing: S&P A-Straps � For flexural reinforcement or (Aramid fibre straps) wrapping/confinement of building elements S&P Laminates CFK � For flexural reinforcement (chap. 10.8) (carbonfiber composite) In the case of prestressed S&P Laminates CFK the C fibres are protected during prestressing and bonding by the cured epoxy resin matrix. As a result, the forces are evenly transmitted into the fibre bundles. With the dry aramid fibres an even force trans-fer from the locking devices of the prestressing system into the fibre bundles is not pos-sible. During the prestressing process the fibres in the centre of the aramid fibre bundles exhibit greater elongation than the external fibres. For this reason the S&P prestressing system uses twisted aramid fibre bundles. The aramid straps have a theoretical fibre cross section of 13.3 mm2/strap. The straps are typically prestressed at a tensile strength of 1’000-1’500 MPa. Prestressing force per S&P A-Strap = 13.3 mm2 x 1’000 MPa = 13.3 kN/strap For the wrap of structural elements typically 15 S&P A-Straps are used per meter. This corresponds to a prestressing force of 200 kN/m1. Prestressing force = 15 straps x 13.3 kN = 200 kN/m1 of the column (1’000 MPa)

Figure 5: Confinement of columns using S&P A-Straps. (For this system a PCT Patent is pending.)

17

8. Substrate requirements for FRP applications The forces from the FRP are transmitted into the substrate through the modified epoxy matrix. FRP applications require a high quality substrate. The various FRP systems re-quire different substrate conditions. The S&P G-Sheet (glass fibre sheet) can be applied to substrates with a low surface bond strength. Glass fibre sheets are suited for the retrofit of historic buildings and the enhancement of the total carrying capacity of masonry walls. For the bonding of the S&P C-Sheet and the S&P A-Sheet, however, a bond strength of the substrate of >1.0 MPa is necessary. Prefabricated laminates produce a concentrated load transfer into the sub-strate. Thus, the application of the S&P Laminate CFK requires a substrate bond strength of 1.5 MPa. Substrate requirements: Product Substrate bond strength S&P C-Sheet / S&P A-Sheet > 1.0 MPa S&P G-Sheet approx. 0.2 MPa S&P Laminates CFK > 1.5 MPa In order to transmit the forces from the S&P FRP system into the substrate through ef-fective bond, surface roughening by sand blasting or grinding is required.

Figure 6: Testing of the bond (tensile) strength

of the bearing substrate.

Figure 7: Roughening of the surface by sand-

blasting or grinding.

18

9. Bonding and anchoring of FRP systems 9.1 Confinement of columns In the case of wrapping structural elements bonding in the lap area of the FRP is carried out as a hoop directional confinement. The load carrying capacity of the substrate is therefore less important. Large scale tests were conducted at TU Gent (Belgium) with wrapped columns where only the FRP lap zone was bonded with epoxy resin. Tests on the axially loaded columns showed identical results to those systems with full-surface bonding and bonding in the FRP lap zone only. Bond strengths between FRP layers Epoxy adhesive (S&P Resin Epoxy 50/55) 15 MPa S&P Resicem 12 MPa Design bond strength 10 MPa

Figure 8: Axial compressive tests

at TU Gent (Belgium) The required anchoring lengths are deduced from the necessary bonding area. Example: S&P C-Sheet 240 (300 gm/m2) Tensile force at width 1’000 mm and limiting design strain 0.6%, see chap. 4.2 � 211kN Theoretical anchoring length = 211 (kN) __ = 21 mm 10 (MPa) x 1’000 (mm) S&P recommends the following lap lengths: S&P C-Sheet 150 mm in the fibre direction S&P A-Sheet 120 mm in the fibre direction S&P G-Sheet 100 mm in the fibre direction 9.2 Design for flexural strengthening using S&P Laminates CFK The bonding of an additional external S&P Laminate CFK on to the tensile stress zone of a structural element subject to bending is carried out with a system approved epoxy resin. Thus, a reinforced concrete structure is produced with an elastic-plastic (steel re-inforcement) and a perfectly elastic tensile element (S&P Laminate CFK). Models for the calculation of the flexural capacity of the composite structure and of the anchor lengths were found by means of bond tests.

19

Figure 9: Related bond failure forces of all bond tests on CFK laminates, depending on the anchor

length (TU Braunschweig, Germany)

The maximum bond strength is obtained at an anchoring length of the CFK laminate of 300 mm. Longer anchoring lengths do not contribute to a further increase in bond strength. The maximum bond strength of a surface applied CFK laminate of width 80 mm is approx. 35-40 kN. The traditional calculation model for the adhesive bonding of FRP reinforcement to con-crete is based on a non-linear brittle fracture mechanics and can be used for any elastic laminate material. The applicability of the traditional models was established by means of bond tests carried out to obtain the General German Approval for S&P Laminates CFK. The bond strengths are transmitted into the substrate between two flexural or shear cracks. Verification of the anchoring of the theoretical laminate end (Point E) is es-tablished if the progression of the bending moment curve is positive. The residual tensile strength of the laminate at Point E corresponds to the theoretical bond failure strength of the laminate. In the case of a negative progression of the bending moment curve (mo-ments at support) the laminate is anchored at the point of zero moment, at the required displacement level and anchoring length.

* for

mor

e de

tails

ask

for t

he E

xper

t Opi

nion

by

Pro

f. R

osta

sy, T

U B

raun

schw

eig

20

Figure 10: Load exposures of the bonding area

- Flexural cracks in the area M with high bend-

ing moments require a confinement rein-forcement to enable the tensile load transfer into the concrete. The concrete thus helps reducing the bending moments.

- In the area Q with varying bending moments

the stresses resulting from the tensile force variation are added to the above mentioned confinement reinforcement.

- In the area V of the end anchoring of the ex-

ternal steel strap binder the tensile strength of the laminate is connected to the concrete and transmitted into the internal reinforcement.

9.3 Anchoring at the ends of the laminates

The S&P design software automatically calculates the required anchor length. In case of anchoring the laminates in a compression zone, no mechanical aid is needed. In the case of steep moment curves, under tension zone, the anchor length of the S&P Laminate CFK as indicated by the S&P design software is often insufficient. In this situation, two alternatives for anchoring at the laminate ends are available.

A) Strengthening of slabs

Figure 11: Anchoring plate

The forces are transferred from the laminate into the substrate through a stainless steel plate fixed with 4 - 6 steel bolts. For design purposes a bond strength of 120 kN is assumed for the anchored CFK laminate with a width of 80 mm. (See bond tests on prestressed S&P Laminates CFK in chap. 9.4.1)

B) Strengthening of beams The ends of the S&P Laminate CFK are wrapped with the S&P C-Sheet 640. Thus, the forces are transferred into the web of the beam. Tests on bending beams, with and without confinement reinforcement of the laminate ends, have been carried out at the University of Lisbon.

21

Testing arrangement

Figure 12: Test at the University Lisbon (P)

Note: Confinement of the laminate end in this case serves to guarantee the required end an-choring. In addition, the FRP confinement improves the shear capacity of the cross sec-tion (see chap. 10.7). 9.4 Prestressed FRP systems 9.4.1 Prestressed S&P Laminates CFK

Figure 13: Bond tests on end anchored S&P

Laminates CFK

At HTA University of Fribourg (Switzer-land) more than 50 bond tests were con-ducted on S&P Laminates CFK with end anchoring using steel plates. Failure (debonding of the CFK laminate under the steel plate) occurred at a prestressing force of 154 kN. For design purposes S&P recommends a maximum tensile strength of 120 kN for the end anchoring with the S&P steel plate. The maximum prestressing force of the S&P Laminate CFK should thus not exceed 120 kN.

The test results show that due to the confinement reinforcement of the beam at the ends of the lami-nates, the bending moment can be increased by approx. 20%. F1: without confinement reinforcement

F3: with confinement reinforcement

22

Maximum prestressing force of the S&P Laminate 150/2000 Type 50/1.2 Prestressing force 0.6% 59.4 kN Type 80/1.2 Prestressing force 0.6% 95.0 kN Type 100/1.2 Prestressing force 0.6% 118.8 kN 9.4.2 Prestressed S&P A-Straps Axial load enhancement

Figure 14: Test at HTA University of Fribourg

The hoop directional confinement in the lap area of the S&P A-Straps is ensured by the application of a bi-directional laminate. Tests show that the confinement is maintained up to failure of the prestressed strap.

Prestressing force: 15 S&P A-Straps/m1 at 13.3 kN = 200 kN/m1 Flexural tests on concrete slabs strengthened with prestressed S&P A-Straps will be conducted in the near future at FH University of Stuttgard (Germany).

23

9.5 Slot application of S&P Laminates CFK The S&P Laminate CFK 10/1.4 with a width of 10 mm and a thickness of 1.4 mm is spe-cially designed to be bonded into slots in concrete or timber structures. A concrete saw is used to cut slots approx. 3 mm wide and 10-15 mm deep into the substrate. The slots are filled with the system approved epoxy adhesive, and the S&P Laminates are pressed into the adhesive.

The performance of slot-applied laminates has been tested at the Technical University in Munich. The test results positively proved that a good and uniform bond exists between the laminate and the concrete. Furthermore, the high tensile strength of the laminate fi-bres was fully utilized prior to shear failure between laminate and surface.

Tests: Comparison of slot-applied and surface-applied CFK laminates by means of bond

strength tests.

Figure 16: Results from the tests

Bond tests :

Figure 15: Testing arrangemant

24

10. Selection of the S&P FRP system Low modulus sheets made from glass or aramid fibres are used to increase the ductility of a structural element. As the raw material price of glass fibres is much lower than of A fibres, glass fibre products (S&P G-Sheet 90/10, Types E or AR) are normally used. Aramid sheets (S&P A-Sheet 120) are ideally applied as impact protection. The ductility enhancement of the different FRP systems has been verified in push-pull tests. Two FRP wrapped columns were compared to a reference column:

• S&P C-Sheet 240 (E-modulus 240’000 MPa) � 1.0 kg in the hoop direction • S&P G-Sheet 90/10 (E-modulus 65’000 MPa) � 3.6 kg in the hoop direction

As the modulus of elasticity of the G fibre is only approx. 25% of the modulus of the C fi-bre, the weight of the S&P G-Sheet applied in the test was raised by a factor of four (in the hoop direction).

Figure 17: Pull-Push Test

Figure 18: Results from a push-pull test

The test results clearly show the more ductile behaviour of the column wrapped with 4 kg of G fibres in the hoop direction compared to a column confined with 1 kg of C fi-bres. This is why structural elements in seismically endangered areas are preferably ret-rofitted with G fibres (see chap. 11). At the Technical University of Gent (Belgium) large scale tests were carried out on circu-lar columns of height 2.0 m and diameter 400 mm, to which different FRP systems had been applied. FRP systems:

• S&P C-Sheet 240 (stretched fibres) • S&P C-Sheet 640 (stretched fibres) • Carbon/glass hybrid (undulating fibres) •

Glass sheet (undulating fibres)

The increase in axial load capacity obtained from the FRP confinement was measured.

25

In order to achieve an identical increase in axial load capacity, the following fibre quanti-ties were required in the confinement direction: • S&P C-Sheet 240 (stretched fibres) 1.0 kg of C fibres • S&P C-Sheet 640 (stretched fibres) 1.6 kg of C fibres • Carbon/glass hybrid (undulating fibres) 2 - 3 kg of fibres • Glass sheet (undulating fibres) ≈4.0 kg of G fibres

Figure 19: Results from the tests at the Techn. Universität Gent (B)

Figure 20: Point of failure

Wrapping with the S&P C-Sheet 240 is suited for the axial load enhancement of circular columns. With 1 kg of C 240 fibres, applied in the hoop direction, identical values were obtained as with 4 kg of G fibres. Large scale test on a column with confinement reinforcement of 5 layers of S&P C-Sheet 240

Figure 21: Results from the tests at the Techn.Universität Gent (B)

In the test the FRP confinement provided an increase in axial load capacity of 57%. In the failure condition the reinforced column showed an axial deformation of 11 mm/m. Under service loading this axial deformation is unfavourable. Thus, the C fibre rein-forcement is used at the stress level of the failure condition in order to provide a suitable safety factor. A maximum reinforcing factor of 1.8-2.0 is therefore realistic.

26

10.1 Axial load confinement of columns 10.1.1 General Strengthening of structural members using FRP wrapping basically leads to a multi-axial compressive stress condition due to the transverse confinement of the concrete. This produces an enhancement of the load capacity of the column. The theoretical design principles used below for the axial load confinement of columns were elaborated by Wang Yung-Chih at the Canterbury University in New Zealand in 1996-2000. His investigations are partly based on test results obtained by Priestley/Seible at the University of San Diego in U.S. A comparison shows that test re-sults as well as design models are identical with the results and models obtained at TU Gent (Belgium).

Figure 22: Dual confinement effect on a rectangular column with an FRP jacket and internal steel hoops

10.1.2 Evaluation of the axial compressive load - axial deformation response Figure 22 shows a cross section of a reinforced concrete rectangular column that is con-fined by an FRP jacket. The eccentric compressive load carried by a short reinforced concrete column, N, is given by N = Nc + Ns (10.1) where Ns = ƒs As (10.2) and Nc = Nc0 + Ncc,ƒ + Ncc,ƒs

= ƒc0 Acu + ƒcc,ƒ Acƒ + ƒcc,ƒs Ac,ƒs (10.3) where Nc and Ns are the compressive loads carried by the concrete and the longitudinal reinforcing bars, respectively. As is the area of longitudinal reinforcement and ƒs is the compressive stress in the longitudinal reinforcement.

27

The concentric compressive strength Nn is, from equation (10.1): Nn = Ncn + Nsn (10.4) Where Ncn and Nsn are the nominal compressive strengths carried by the concrete and the longitudinal reinforcing steel bars, respectively, when the axial compression reaches the ultimate state of the column. Wang et al assume that the ultimate limit state in a concentrically loaded column is associated with 1% axial strain. With Poisson’s ratio conservatively assumed to equal v = 0.5 at this strain level, the transverse strain at 1% axial strain is equal to 0.5%. Wang et al have assumed the reinforcing steel behaves as an elasto-plastic material. The nominal compressive strength carried by the concrete, Ncn, results from the stresses in three distinct regions shown in Figure 22. At 1% axial strain the unconfined concrete has reached its peak strength, ƒ’c , and has degraded to a residual strength to 0.3 ƒ’c. From Eqs, 10.2 & 10.3 with ƒc0 = 0.3 ƒ’c Nsn = ƒsy As (10.5) Ncn = 0.3 ƒ’c Acu + ƒ’cc,ƒ Acƒ + ƒ’cc,ƒs Ac,ƒs (10.6) Where ƒ’cc,ƒ and ƒ’cc,ƒs are the confined concrete compressive strengths due to the single confinement of the external jacket and the dual confinement of the jacket and steel hoops, respectively. Acu, Acƒ, and Ac,ƒs are the confined areas with respect to different confining regions. ƒsy is the yield strength of longitudinal reinforcement and As is the area of longitudinal reinforcement. For design purposes it is necessary to reduce the nominal concentric strength given in Eq. 10.4, to account for variations in the materials properties, scatter in the design equa-tion, bending of the columns, nature and consequences of failure and reduction in load carrying capacity under long-term loads. This reduction results in a dependable concen-tric strength, Nn, for short columns given by,

Nnd = ms

sncnk

mc

NN67.0γγ

+ (10.7a)

or Nnd = ms

sncnc

mc

NN788.0γγ

+ (10.7b)

where γmc and γms are the partial safety factors for concrete and steel respectively, Ncnk and Ncnc are the compressive forces carried by the concrete if derived from the cube or cylinder strengths respectively and Nsn is the compressive force carried by the reinforc-ing steel.

28

Using γmc = 1.5 and γms = 1.15, equation 10.7 can be written as: Nnd = 0.447 Ncnk + 0.87 Nsn (10.8a) or Nnd = 0.525 Ncnc + 0.87 Nsn (10.8b) Therefore, Eq. 10.8 becomes, Nnd ≥ N* (10.9) Where N* is the design concentric axial load in the column. The compressive load carried by the concrete, Nc, results from the loads sustained by three distinct regions. In Eq. 10.3, Nco is the load carried by the unconfined concrete re-gion, Acu, and ƒco is the compressive stress of unconfined concrete. Ncc,ƒ is the load car-ried by the effective area of concrete confined by the FRP jacket, Ac,ƒ and ƒcc,ƒ is the confined area and compressive stress of concrete confined by the FRP jacket, respec-tively. Ncc,ƒs is the load carried by the effective area of concrete Ac,ƒs confined by both the FRP jacket and the steel hoops, and ƒcc,ƒs is the corresponding stress. Hence, the entire uni-axial stress-strain relationship for a concentrically loaded column wrapped with an FRP jacket can be obtained if the constitutive stress-strain relationships for each of the regions and for the reinforcing steel are known. The area of effective confining core confined by the steel hoops can be ignored in most practical applications. The area of effective confining core for the FRP jacket is given by the following expression: Acu = Acc,ƒ - Ae,ƒ (10.10) Where Acc,ƒ is the area of concrete confined by the jacket and Ae,ƒ is the area of con-crete effectively confined by the jacket. In the case of a rectangular column, the areas, Acc,ƒ and Ae,ƒ are given by, Acc,ƒ = txty – As – (4 - π) r2 (10.11)

Ae,ƒ = txty – 2s

2y

2

π)r(4Atan3

w'w'−−−

+ ƒƒ θx (10.12)

Figure 22 defines the variables used in above equations. The area of concrete effec-tively confined by the jacket, Ae,ƒ , is limited by w’ƒx < 2 w’ƒy when the longer side is w’ƒx or by w’ƒy < 2 w’ƒx when the longer side is w’ƒy .

29

In the case of a circular column, see Figure 23,

Acc,ƒ = Ae,ƒ = s

2

A4D π − (10.13)

Where D is overall column diameter. It can be seen that Acc,ƒ = Ae,ƒ makes Acu = 0 and consequently the confined effectiveness of circular column is greater than that for a rec-tangular column. 10.1.3 Determination of the compressive strength of the confined concrete The compressive strength of confined concrete, as proposed by Mander et al, ƒ’cc is given by: ƒ’cc = kc ƒ’c (10.14) In which ƒ’c is the cylinder concrete strength and kc is the concrete strength enhance-ment factor. Factor kc depends on the bi-axial state of stresses induced by the lateral confining pressures. This factor is expressed as: kc = α1 α2 (10.15) Where α1 is a strength enhancement factor that considers the concrete to be subjected to a tri-axial stress state with bi-equal confining stresses and α2 is a reduction factor that considers any deviation from bi-equal confining stress concept.

Figure 23: Dual confinement effect on a

circular column with an FRP jacket and internal steel hoops

30

Mander further proposed that:

α1 =

ƒ−

ƒ+ 1

'F

1.6'

F 7.9411.8 1.25

c

l

c

l (10.16)

and that

α2 = 1'

F0.8F

0.6 F

1.4 c

l2

l

l

l

l +ƒ

ƒ−

ƒ (10.17)

In the above equations, Fl and ƒl are the maximum and minimum confining lateral stresses, respectively. 10.1.4 Evaluation of the lateral confining pressure To calculate the concrete strength enhancement factors, the lateral confining pressure must be found. The evaluation of the lateral confining pressure due to an elastic jacket and internal reinforcing steel hoops for rectangular and circular columns is derived be-low. Confinement provided by the FRP jacket Rectangular columns The lateral confining stresses induced by an FRP jacket in the x and y directions, ƒl,ƒx and ƒl,ƒy, are, ƒl,ƒx = ρƒxƒƒ (10.18) ƒl,ƒy = ρƒyƒƒ (10.19) Where ƒƒ is the stress in the FRP jacket. The reinforcement ratios ρƒx and ρƒy are defined as,

ρƒx = yt

t2 ƒ

(10.20)

ρƒy = xt

t2 ƒ

(10.21)

where tƒ is the nominal jacket thickness and tx and ty are the overall column cross sec-tion dimensions.

31

The stress in the jacket can be found from transverse strain compatibility. Assuming v = 0.5 and for 1% axial strain, the transverse strain is 0.5%, hence the stress in the jacket is: ƒƒ = 0.005Eƒ / γmE (10.22) Where Eƒ is the modulus of elasticity of the FRP material and γmE is the partial safety factor for the modulus of elasticity. S&P recommends: γmE: for S&P C-Sheet 240 (uni-directional): 1.1 – 1.2 γmE: for S&P G-Sheet 90/10 (bi-directional): 1.5 – 1.8 Circular columns The lateral confining stress induced by the jacket, ƒl,ƒ is: ƒl,ƒ = ρƒƒƒ (10.23) where ƒƒ is the stress in the jacket determined from equation 10.22. The confinement re-inforcement ratio ρƒ is defined as,

ρƒ = 4 Dt ƒ (10.24)

where tƒ is the nominal jacket thickness and D is the overall column diameter. 10.1.5 Design example A 300 mm by 450 mm rectangular column is reinforced with 6 bars with As = 1’884 mm2. The concrete cube strength ƒcc = 22.4 MPa and the yield strength of the reinforcing bars ƒy = 400 MPa. Determine the axial load carrying capacity if the column is to be jacketed with 4 wraps of S&P C-Sheet 240 (300 gm/m2). Solution A: Determine the design properties of the S&P C-Sheet 240 (300 gm/m2) • Design ultimate strain: εud = εult / γmF and γmF = γmƒ x γmm = 1.4 x 1.4 = 1.96 (according recommendation UK 55 Guideline) εud = 1.55% / 1.96 = 0.79% (E.1) ⇒ εud > 0.5% as required by method proposed by Wang et al.

32

• Design elastic modulus: Eƒd = Eƒ / γE = 240’000 / 1.1 = 218.2 Mpa (E.2) B: Determine the distinct unconfined and effectively concrete areas.

Assume the corners will be rounded to a 50 mm radius. Hence from equations 10.10 to 10.12 ω’ƒx = (300 – 2 x 50) = 200 mm ω’ƒy = (450 – 2 x 50) = 350 mm Acc,ƒ = 300 x 450 – 1’884 – (4 - π) 502 = 130’969.9 mm2 (E.3)

Ae,ƒ = 300 x 450 – 3

350200 22 + tan 45° - 1’884 – (4 - π) 502 (E.4)

= 76’803.3 mm2

Acu = 130’969.9 – 76’803.3 = 54’166.6 mm2

C: Determine the confined concrete properties ƒ’c = 0.85 ƒck = 0.85 x 22.4 = 19.0 Mpa (E.5) Now, find the lateral confining stresses from equations 10.18 to 10.22 ƒƒ = 0.005 x 218’182 = 1’091 Mpa (E.6) tƒ = 4 x 0.176 mm = 0.704 mm (E.7)

ρƒx = 4500.704 x 2 = 0.31% (E.8)

ρƒy = 3000.704 x 2 = 0.47% (E.9)

ƒl,ƒx = 0.0031 x 1’091 = 3.41 Mpa (E.10) ƒl,ƒy = 0.0047 x 1’091 = 5.13 Mpa (E.11) Fl = max(5.13 , 3.41) = 5.13 Mpa (E.12) ƒl = min(5.13 , 3.41) = 3.41 Mpa (E.13)

33

D: Determine the concrete enhancement strength ratio, kc from equations 10.15 to 10.17:

c

l

'Fƒ

= 5.13 / 19 = 0.27 (E.14)

c

l

'ƒƒ = 3.41 / 19 = 0.18 (E.15)

l

l

Fƒ = 0.665 (E.16)

α1 = 1) 0.27 x 1.60.27 x 7.941 1.8 ( 1.25 −−+ = 2.20 (E.17) α2 = [1.4 x 0.665 - (0.665)2 – 0.8] 10.27 + = 0.93 (E.18) kc = 2.20 x 0.93 = 2.05 (E.19) Consequently, the confined concrete strength is, from equation 10.14 and E.5 ƒ’cc = 2.05 x 19.0 = 39.0 Mpa (E.20) E: Ultimate load of jacketed column: Now the axial compressive load carried by the concrete is, from equations 10.6, Ncnc = 0.3 x 19 x 54’167 + 39 x 76’803 = 3’304 kN (E.21) and the axial compressive load carried by the reinforcing steel is, from equation 10.5, Nsn = 400 x 1’884 = 753.6 kN (E.22) The ultimate load is, from equation 10.8b, Nnd = 0.525 x 3’304 + 0.87 x 753.6 = 2’390 kN (E.23) F: Ultimate load of “as-built” column: Nnd = 0.525 [(300 x 450 – 1’884) x 19] + 0.87 (1’884 x 400) (E.24) = 1’983 kN Thus the application of the jacket will result in an ultimate axial load enhancement of 407 kN or 21% of the “as-built” column ultimate load. The FRP confinement will guaranty the required safety factor at the ultimate state level. The FRP will not be activated at service level it means, that there is only a small axial deformation in the column at ser-vice level.

34

10.2 Prestressing of axially loaded columns In order to improve the compressive strength of a rectangular column without large al-lowable axial compressive strains, the A fibre should be prestressed. Prestressed S&P A-Straps are suited for this field of application. The external prestressing produces a de-fined hoop effect. Due to prestressing, cracks in the existing column are closed and premature breaking of the fibres on the edges of the column is likewise prevented. Thus, increasing in com-pressive strength can be carried out at a minimal axial strain level. Prestressed S&P A-Straps are suited for the enhancement of the serviceability of circular and rectangular columns. Due to prestressing higher reinforcing levels can be realised. The new reinforcing concept is confirmed by preliminary test results obtained in co-operation with HTA University of Fribourg: Axial compressive strength tested on unreinforced low cross section columns at HTA University of Fribourg (Switzerland) Column dimensions: Side length = 20/20 cm Height = 65 cm Rounding of edges = 2.5 cm

Figure 24: Reference column Failure at 1’353kN Figure 25: Wrapped column (prestressing of S&P A-Strap at 200 kN per m1 of column height) Failure at 1’718 kN

35

Figure 26: Test conducted at HTA University of

in Fribourg (Switzerland)

Interpretation of test: The compressive strength of the rein-forced square column could be increased by 27%, compared to the reference col-umn, without producing large axial strains. Additional research works on the new strengthening concept will be carried out in 2002.

10.3 Combined moment and axial load on columns In the case of both moments and axial load applied to a circular or rectangular column, two types of reinforcing are applied. The additional axial force is absorbed by the FRP wrap (chap. 10.1/10.2). The additional moment is absorbed by a slot- or surface-applied laminate that is placed prior to the FRP confinement. The S&P Laminate CFK is dimen-sioned for the additional moment effect using the S&P design software. 10.4 Replacement of corroded reinforcement using FRP In traditional concrete repair, the corroded steel reinforcement may be replaced by FRP. Thus, the original safety factor of the constructional element is guaranteed after the re-pair.

Concept:

- Remove all spalled concrete. - Coating of the cleaned steel reinforcement with S&P Resicem (cementitious ep-

oxy resin matrix). - Conventional concrete repair using a cementitious mortar (S&P Repcem). - New: The corroded reinforcement may be replaced with FRP. - Final coating of the element (cementitious plaster or vapour permeable coating,

such as acrylate or PU based coating).

36

Replacement of corroded flexural reinforcement with S&P Laminates CFK bonded into slots.

Figure 27: Column with a slot-applied laminates

Note: With the slot-applied CFK laminate higher forces can be transmitted into the substrate (see chap. 9.5). Thus, the S&P Laminate CFK can be utilised to a higher degree (see chap. 10.6.3). For design purposes a tensile strength of 2’000 MPa is assumed for the slot-applied laminate. This means that approx. 25% of the surface of the corroded inter-nal steel (yield: 500 MPa) are replaced by the slot-applied CFK laminates. The corroded reinforcing stirrups are replaced by confinement reinforcement with S&P C-Sheets 240.

Figure 28/29: Column with S&P C-Sheet

confinement reinforcement. A = distance between the S&P C-Sheet 240

A ~ 0.8 x diameter of round column A ~ 0.8 x side length(short side) of rectangular column

Note: The S&P C-Sheet 240 is applied to replace the corroded stirrup reinforcement. The maximum spacing between the C-Sheet shall not exceed the internal lever arm of the column (analogy to strut & tie model).

37

10.5 Seismic upgrading of columns For the seismic upgrading of columns different FRP systems are used. While in the western part of the Unites States embrace reinforcement with E-glass (S&P G-Sheet 90/10, Type E) is commonly used, in the seismically endangered zones of East and Middle Europe the alkali-resistant AR-glass (S&P G-Sheet 90/10, Type AR) in combina-tion with the water vapour permeable "Resicem" matrix is usually applied. Aramid fibres (S&P A-Sheet 120) are only used in exceptions because of their high price. At the Technical University in Porto (Portugal) an investigation into the seismic resis-tance of carbon fibre based FRP systems was conducted in a series of push-pull tests. To this end, square columns with a side length of 200 mm were reinforced with FRP. S&P Laminates CFK were bonded into slots to improve the longitudinal flexural strength. Partial wrapping was done with the S&P C-Sheet 240. The tests indicated that the maximum displacement angle of 22 mm/m (maximum loading of the testing device) was reached.

Figure 30: Tests at the Technical University in Porto (Portugal)

The results show that the combined application (into slots and as wrapping) of the high modulus C fibre likewise produces a high reinforcing effect under seismic exposure. (Note: see chap. 11) 10.6 Flexural strengthening with non-prestressed FRP systems In addition to the prestressed systems, three non-prestressed FRP systems are avail-able for flexural strengthening:

• Surface bonded S&P Laminates CFK • S&P C-Sheet 240 (uni-directional C fibre sheet) • Slot-applied S&P Laminates CFK

Vertical load

Horizontal displacement

Axial confinement using slot-applied S&P Laminates

Axial load enha-cement using S&P C-Sheet 240

38

10.6.1 Surface bonded S&P Laminates CFK During tests conducted for the General Approval of the S&P Laminate CFK in France, Germany, UK, Korea etc. the bond of the laminate on beams was investigated. Flexural tests on beams conducted at TU in Braunschweig (Germany) showed that debonding of the laminate from the substrate depends on the strain in the laminate and the plastic strain in the reinforcing steel. Debonding of the reinforcing steel initiated at an elongation of the laminate of approx. 0.65%; this corresponds to 5.7 times the limiting strain of the internal reinforcement. Failure of the beam occurred at an elongation of the laminate of 1.3%. Bending tests on slender slabs conducted at HTA University of Fribourg showed that premature debonding (partly at a limiting strain of 0.6-0.7%) of the CFK laminate is pos-sible. For this reason, a limiting design strain for CFK laminates of ε = 0.6-0.8% has been defined in internationally recognised Guidelines (UK Concrete Society Report TR55 and ACI 440) as well as in the General Approvals for France, Germany etc.

Limiting design strain of non-prestressed CFK laminates: 0.6 – 0.8% (for flexural strengthening)

The S&P design software is available for different national code systems (ACI, BS, French Standard, Eurocode 2, DIN 1045 old and new). In the Eurocode 2 version the limiting design strain values for the S&P Laminate CFK as prescribed in the General German Approval are used. 10.6.2 S&P C-Sheet 240 In cases of substrates with a low surface bond strength (<1.5 MPa) or with minimal laminate requirements, it is possible to replace the S&P Laminate 200/2000 by the S&P C-Sheet 240. With the use of the S&P C-Sheet 240 for flexural strengthening the sur-face for the load transfer into the substrate is raised by a factor of 5 to 10. This allows the product to be applied to surfaces with a bond strength of 1.0 MPa. Dimensioning of the S&P C-Sheet 240 for flexural reinforcement is carried out using the S&P design software. The software calculates the required cross sections of the S&P Laminate 200/2000. The theoretical E-modulus of the fibre of the S&P C-Sheet 240, due to the reduction by the safety factor of 1.2, corresponds to the modulus of the S&P Laminate 200/2000. The calculated laminate cross sections can thus be converted into the theoretical fibre cross sections of the S&P C-Sheet 240. Example: According to the software 0.4 cm2/m of S&P Laminate CFK 200/2000 are required. Instead of placing S&P Laminates CFK (50/1.4 = 0.7 cm2) at a spacing of 1.75 m = 0.4 cm2/m, S&P C-Sheet 240 (200 gm/m2) with a width of 300 mm and a theo-retical fibre cross section of 0.117 mm x 300 mm = 35.10 mm2/layer are applied at a spacing of 88 cm = 0.4 cm2/m.

39

Since the minimal laminate spacing (a) according to the General Approval is limited as follows, a <<<< 20% of the width a <<<< 5 x the thickness of the concrete slab to be reinforced a conversion into the theoretical C-Sheet cross section is often advisable. In the above example a laminate spacing of 1.75 m is required. This is not possible for a concrete slab of 20 cm. The spacing of the S&P C-Sheet 240, however, fulfils the requirement: a < 5 x 20 cm = 1.0 m a actual = 0.88 m For the verification of the anchoring at the end of the C-Sheet the larger area available for the load transfer may be taken into account. Thus, verification of the anchoring is a simple task. 10.6.3 Slot-applied S&P Laminates CFK Three point load tests with a span of 2.5 m were carried out on various reinforced con-crete beams.

Figure 31: Test arragement

Each sample was either reinforced by a surface-applied CFK laminate 50/1.2 or by two slot-applied CFK laminates 25/1.2. - On test beams A1 and B1 failure occurred due to debonding of the CFK laminate. - On test beam A2 failure occurred due to tensile fracture of the slot-applied laminate. - On test beam B2, with a low shear reinforcement made of steel, shear failure oc-

curred in the concrete.

40

Interpretation of results from test beams A At equal stiffnesses, the ultimate load was more than doubled using the slot-applied laminate. This is due to the high utilization of the tensile strength of the CFK laminate.

Figure 32: The load-deflection curves of test beams A1 and A2

Interpretation of results from test beams B The load-deflection curves are almost iden-tical except that the slot-applied CFK lami-nate exhibited a substantially higher ulti-mate load.

Figure 33: The load-deflection curves of test beams B1 and B2

Benefits of slot-applied laminates

• The improved utilization of the laminate permits higher loads to be applied and laminate cross sections to be reduced.

• The quality of the substrate (tensile strength of the surface) is less important. Slot-applied laminates can also transfer loads into substrates with a low bearing capacity (brickwork, masonry).

The slot application is more economical than levelling and roughening required for surface-applied laminates.

• The slot-applied laminate is protected against mechanical damage. Better per-formance is achieved in the event of a fire, thus reducing the cost of fire protec-tion measures.

Ideal fields of application of slot-applied laminates are:

Strengthening of the negative moment (moment at support) • Slot-application to flexural elements (flexural strengthening)

Dimensioning of slot-applied S&P laminates can likewise be conducted using the S&P design software. The software calculates the required laminate cross sections of surface bonded S&P Laminates CFK 150/2000, e.g. for the strengthening of the moment at sup-port. Since the slot-applied laminates can be utilised to a higher degree, the laminate cross section that would be required for surface bonded laminates is in practice reduced

41

by 25-30% for the slot-applied laminates. Bond tests on slot-applied laminates (chap. 9.5) show that the verification of the anchoring is a simple task. Debonding of the slot-applied laminate does not occur. On a slot length of 20-30 cm the ultimate tensile strength of a slot-applied laminate of width 10-50 mm and thickness 1.4 mm can be transferred into the concrete substrate. 10.7 External shear reinforcement Flexural strengthening with S&P Laminates CFK often requires an increase in shear strength. This can be achieved by bonding of S&P C-Sheets 640 around the web of the beam.

The shear force VSdf in the strengthened condition is shared between the internal steel stirrups and the bonded S&P C-Sheet 640.

The design software distinguishes two cases:

Case 1: VSdf � VRd3 internal resistance The internal shear reinforcing is not sufficient to absorb the shear force in the strengthened condition VSdf

∆V = VSdf – VRd3

∆V: Differential shear force (portion of the shear force VSdf that is absorbed by the addi-tional FRP shear reinforcing)

or

η1)(η∆V −= VSdf η : flexural reinforcing level

In this case, the S&P C-Sheet 640 is designed to fulfil the requirements of the as-sociated strut & tie model and must be anchored in the compression zone.

42

Figure 34: Slots are cut into the concrete slab to allow anchoring and wrapping of the S&P C-Sheet 640 in the compression zone.

Figure 35: The S&P C-Sheet 640 is anchored in to

the web of the bending beam, 150 mm above the neutral axis. Hoop directional confinement of the S&P C-Sheet is re-quired (DIN) depending on the size of the transverse load

Case 2: The internal shear reinforcement is sufficient to absorb the transverse

force in the reinforced condition VSdf.

VSdf ≤ VRd3 SdfVVη

η )1( −=∆

In this case, the S&P C-Sheet 640 is statically not necessary and a minimum strengthening is dimensioned depending on the flexural reinforcing level. Design is done using the „FRP Lamella“ software.

Shear capacity and its strengthening according to EC 2

EC 2 limits the transverse force VSd which can be transmitted by the concrete only or in combination with the shear reinforcement. This shear capacity is designated by the de-sign values VRd1 to VRd3.

It is assumed that the relevant formulas from EC 2 are known.

VRd1 Limiting shear force where the element has no shear reinforcement. The shear force is carried by the concrete only.

VRd2 Maximum limiting shear force. Relevant to the shear capacity is the bear-ing capacity of the inclined compression struts.

VRd3 Limiting shear force with shear reinforcement. The shear force is transmit-ted by the concrete and the shear reinforcement.

The lower design value VRd1 is of relevance in slabs that are typically retrofitted without shear reinforcement.

43

Dimensioning of the additional shear reinforcement Shear reinforcements can best be carried out using high modulus carbon fibre sheets (S&P C-Sheet 640). These uni-directional sheets with a modulus of elasticity of approx. 640’000 MPa are much easier to handle than steel reinforcement and therefore are eco-nomical to use notwithstanding the high material costs. The level of required shear reinforcement, with externally bonded steel strap binders or sheets, is designed from the differential force ∆V. As the internal reinforcement and the additional external shear strengthening must be considered as parallel elastic-plastic elements, uniformity of strain must be maintained in the reinforced condition. Thus, the elongation of the stirrups is limited to εlimit = 2 ‰. This limiting value is recommended when using carbon fibre sheets in such an application. Dimensioning of the external reinforcement with steel strap binders or sheets is based on the EC 2 standard procedure

fdferfw, d0,9

VAσ⋅⋅

∆=

where for the internal lever arm the approximate value of 0.9⋅df (useful height of the laminate) may be used. For carbon fibre sheets the limiting stress must be reduced:

fdlimitfd E⋅=εσ

Carbon fibre sheets possess no inherent stiffness and, when used as shear reinforce-ment, they are manually applied on the job site. Thus, it is questionable whether the high E-modulus of 640’000 MPa is actually achieved during application. For design purposes it is therefore recommended that the modulus of elasticity of the carbon fibre sheets be reduced by a safety factor of 1.2.

10.8 Flexural strengthening with prestressed S&P Laminates CFK

Figure 36: Installation of the prestressing system

44

45

46

10.9 Special applications using FRP systems 10.9.1 Repairs to historic buildings Application to masonry

Figure 37: Repairs to a vault

Figure 38: FRP strengthening of a truss S&P Laminate

CFK: as tensile members S&P C-Sheet 240: wrapping of columns S&P C-Sheet 640: wrapping of junction points

Studies of moisture condition in a structure, during Building Physics surveys, must be carried out by an experienced engineer. The selection of G or C fibre systems depends on the quality of the substrate. For bi-directional masonry reinforcements, e.g., the S&P G-Sheet 50/50 is suitable. The glass fibres in each direction of this woven sheet are dis-tributed in the ratio of 50:50. Application to steel Figure 39: Tower Bridge in London

47

Some of the grey iron castings of the Tower Bridge in London exhibited corrosion and cracking and thus required retrofitting. As grey cast iron cannot be welded, reinforcing was carried out with the S&P C-Sheet 240. As a result, the use of prestressed CFK laminates for the retrofit of old composite bridges made of steel girders and concrete has gained acceptance in UK. End anchoring of the CFK laminate to the steel flange is easy to achieve. 10.9.2 Applications in timber construction

Figure 40: Example of an application in timber construction

The fundamental principles for the mechanics of FRP retrofitting of timber structures have been elaborated in two dissertations at Technical Universities in Germany and in a thesis at BOKU in Vienna. The design rules will be established this year in a research carried out at ETH in Zurich. S&P hopes to be able to offer a design software for FRP retrofitted timber elements presumably by the end of the year 2002.

The special S&P Laminate 10/1.4is applied into saw

notches in the load relieved timber beam.

Timber beam

Timber beam

48

γ > 1.0 γ < 1.0

In the event of failure of the CFK laminate, the residual safety factor γγγγ determines the necessary measures required for fire

protection. The S&P design Software automatically outputs the residual safety factor in the case of failure of the

S&P Laminate CFK.

49

11. Seismic retrofitting of reinforced concrete columns and frames using FRP

11.1 Reinforced concrete columns 11.1.1 Strengthening methods for reinforced concrete columns The earthquake resistance of many of the older reinforced concrete columns is too low: Their flexural strength and ductility are not sufficient. On the one hand, these shortcom-ings are due to the fact that the lap splice lengths of the internal reinforcement are too short and, on the other, that the anchoring length of the internal reinforcement of adja-cent structural elements is too short. Typical overlapping in the compression zone, with a length of 20 times the cross section of the reinforcement, is not sufficient to transmit all the post-elastic forces from the reinforcing bars. Under seismic action insufficiently an-chored reinforcing bars show an unacceptable behaviour. When exposed to cyclical bending moments and potential additional tensile loads, these building components are subject to premature failure. This becomes especially critical in the case of foundation connections: If the projecting reinforcement exhibits too little overlapping, under seismic action it may form an undesirable plastic hinge. Unfortunately, this type of deficient con-struction detail is frequently found in older buildings and bridges. Often, also the trans-verse strength of supporting elements is insufficient. A solution to enhance these deficiencies of supporting elements is the confinement. The earthquake resistance of a column with confinement reinforcement on its connections is improved due to the radial stresses. For the confinement (wrapping) of existing struc-tures various methods are available. Three different strengthening methods were com-pared in large-scale tests: • Steel jackets • Concrete casing • FRP confinement The traditional methods with steel and concrete are well known and will therefore not be discussed. 11.1.2 Reinforced concrete columns retrofitted with FRP jackets During the last years the performance of reinforced concrete columns retrofitted with FRP jackets has been verified in several scientific studies. Design concepts for the di-mensioning of FRP confinements are available (e.g. M.J.N. Priestley, F. Seible, G.M. Calvi, "Seismic design and retrofit of bridges"). The confinement of columns with a rectangular cross section is more difficult than that of a column with a circular cross section. The radial stresses are concentrated at the cor-ners. In order to achieve a defined confinement force (radial confinement), S&P has de-veloped a prestressing system. Prestressing produces a defined three-dimensional stress level. In the case of non-prestressed FRP jackets, the three-dimensional stress level is only achieved when transverse strain of the concrete occurs. Prestressing in-creases the reinforcing efficiency of the FRP reinforcement. With this new method suffi-cient radial stress can thus be provided in the case of columns with a rectangular cross section.

50

Unlike steel jackets, FRP reinforced columns do not show any reduction in performance. Additional tests showed that with FRP confinements potential shear failure can be avoided very effectively. Uni-directional confinement (horizontal) of columns with FRP jackets does not affect the vertical stiffness of the element, an essential part of seismic strengthening. Dimensioning of the FRP confinement is basically identical to the dimen-sioning adopted for steel stirrups. The limiting design strain of FRP is assumed as approx. 50% of the breaking elongation of the fibre. The high safety factor guarantees that the shearing mechanism of the reinforced concrete column is not reduced by ex-cessive concrete strains. The FRP confinement acts in two ways: • Under tensile loading the effect of the FRP confinement on the plastic hinge region of

the column is that the overlapping of the reinforcement, or of the anchoring of the lon-gitudinal reinforcement respectively, is strengthened, and the post-elastic behaviour of the reinforcing bars can thus be fully utilised.

• Under compressive loading of the plastic hinge region the confinement prevents buckling of the longitudinal reinforcement as well as spalling of the concrete cover.

11.1.3 Confinement to improve flexural ductility The primary aim of the confinement is the enhancement of the flexural ductility. Columns with insufficient internal stirrup reinforcement cannot sustain large non-elastic rotations in the plastic hinge region. FRP confinements are suited to increase the flexural ductility of such supporting elements. Tests on circular columns retrofitted with FRP clearly indi-cated that they are able to increase the ductility more effectively than conventional steel jackets. The reason for this is the linear-elastic behaviour of the FRP confinement. Seismic response can cause steel jackets to undergo tangential deformation. On unloading residual plastic strains remain in the jackets. The effectiveness of the steel jacket is therefore reduced for each successive seismic response, and for each new load cycle higher hoop strains are required. Due to the linear-elastic behaviour of FRP there is no cumulative damage. Successive cycles cause similar hoop strains. This im-proved behaviour compared to steel jackets is taken into account in the dimensioning concepts for FRP confinement that have been deducted from testing programmes. Circular columns The maximum concrete compressive strain εcu of the column can be calculated as fol-lows:

'cc

FRPFRPscu f

εf2.5ρ0.004ε += [M.J.N. Priestley, F. Seible, G.M. Calvi]

The ratio of volumes of the confinement sρ of a circular column is defined as follows:

D4tρ FRP

s =

51

FRPf and FRPε are the maximum stress and strain in the FRP confinement. The concrete compressive strength '

ccf of a confined reinforced concrete column is de-fined as follows:

−−+= 1.254

f2f

f7.94f12.254ff '

c

'l

'c

'l'

c'cc [M.J.N. Priestley, F. Seible, G.M. Calvi]

In the formula '

cf represents the characteristic concrete compressive strength and 'lf the

effective lateral confinement stress (stress of the theoretical fibre cross section). From the expressions for εcu and sρ , the required confinement strength tFRP (theoretical fibre thickness of the FRP confinement) can be deducted:

( )FRPFRP

'cccu

FRP εfDf0.004ε0.1t −=

Rectangular columns Push-pull tests on rectangular columns retrofitted with GFRP likewise indicated an in-crease in ductility. Figure 41 shows a rectangular column retrofitted with GFRP after failure. In the test a displacement ductility (failure) of ∆µ = 8 was reached. This corre-sponds to a displacement angle of approx. 4%.

Displacement ductility factor firstyield at ntdisplacemeultimate at ntdisplacemeµ∆ =

M.J.N. Priestley, F. Seible und G.M. Calvi, „Seismic design and retrofit of bridges“

Figure 41: Rectangular column with GFRP confinement

52

In several test series it could be shown that with rectangular cross sections com-pared to circular columns with an identical cross section an increase in ductility of approx. 50% is obtained. The maximum concrete compressive strain εcu of the rectangular column is calculated as follows:

'cc

FRPFRPscu f

εf1.25ρ0.004ε += [M.J.N. Priestley, F. Seible, G.M. Calvi]

The ratio of volumes of the confinement sρ of a rectangular column is defined as fol-lows:

+=bh

hb2tρ FRPs

In the formula b and h represent the dimensions of the cross section. From the expressions for εcu and sρ , the theoretical required FRP confinement thickness can likewise be deducted:

( )

+−=

hbbh

εff0.004ε0.4t

FRPFRP

'cccu

FRP

Confinement area In Table 1 the required confinement areas (l) are indicated. Distinction is made between one-sided or double-sided restraint of the columns. La and Lb represent the distances from the support to the centre of moments. Restraint of the column Cross section of

the column Column length Area

One-sided D L D ≤ l ≤ 0.25 L Double-sided: Ma Mb

D La Lb

D ≤ l ≤ 0.25 La D ≤ l ≤ 0.25 Lb

Table 1: Confinement area

11.1.4 Confinement to improve the lap splices of the internal reinforcement The load transfer from the steel reinforcement into the concrete leads to the formation of micro-cracks in the concrete that will reduce the bond between steel and concrete. Ret-rofitting with FRP jackets, and with prestressed FRP systems in particular, enhances this bond.

53

Dimensioning is carried out as follows:

−=FRP

assbs 0.0015E

fµpl/ fA2ρ

fa: active hoop stress Ab: surface of the longitudinal reinforcement fs: stress in the longitudinal reinforcement µ: coefficient of friction (1.4) p: area of influence of the fractured surface in the lap splice region ls: length of the lap splice The prestressed S&P A-Strap is specially suited for this field of application. Sheets with a low modulus of elasticity, such as sheets made of G fibres, are less suited for these applications. Confinement area If strengthening of the lap splices is the only reason for the FRP confinement, a wrap above the lap splice is of no use. It only becomes necessary if the lap splice is not lo-cated in a plastic hinge region. 11.1.5 Confinement for external shear strengthening FRP confinements, similarly to steel jackets, are very efficient in improving the shear re-sistance of a building component. Since FRP exhibits a linear-elastic behaviour until failure, the design value adopted for external steel reinforcement has to be slightly modi-fied. A limiting design strain of the FRP of 0.2-0.3% has to be used. For this type of ap-plication the S&P C-Sheet 640 with a breaking elongation of 0.4% is ideally suited. Post-reinforcement of the shear resistance can be calculated as follows: Circular columns

θ cot Dft2

V FRPFRPFRPπ= [M.J.N. Priestley, F. Seible, G.M. Calvi]

D: cross section of the column θ : 35°

Rectangular columns

θcot h 2 FRPFRPftVFRP = h: column dimension in parallel to the shear stress θ : 35°

54

Confinement area In the case of reduced shear resistance, the FRP confinement should be applied to the plastic hinge regions at a height of 2D for circular columns, or 2h for rectangular col-umns respectively. 11.1.6 Flexural strength of the column As a result of the confinement of a column, its flexural stiffness is increased and it is thus subject to higher forces. This is a critical issue and needs careful attention. The in-crease in flexural strength depends on the elected strengthening method, the material parameters and the shape of the columns. Table 2 shows the increase in flexural strength of a column as a function of the elected strengthening method.

Column Reinforcing Steel Concrete FRP Circular Plastic hinge 10 - 20 20 - 50 0 - 5 Shear 20 - 40 25 - 75 0 - 5 Rectangular Plastic hinge 20 - 40 20 - 50 0 – 10 Shear 40 - 70 25 - 75 0 – 5 Table 2: Increase in flexural strength (%)

FRP retrofitting causes a substantially lower increase in flexural strength compared to the application of steel jackets or concrete casing. 11.1.7 Comparison of strengthening methods A comparison of the different strengthening methods is given in Table 3. The results demonstrate the favourable behaviour of FRP confinements compared to conventional procedures. The main benefit of the FRP confinement is the minimal increase in flexural strength, despite its high capacity in improving ductility. As a result, the FRP retrofitted structural element is not subject to additional forces, and premature failure of the frame-work is thus prevented. A defined three-dimensional stress condition is obtained using the prestressed S&P A-Straps. Prestressing provides an increase in concrete compressive strength, without producing large axial compressive strains in the load bearing element. Due to the in-creased compressive strength higher loads can be transmitted. The application of FRP confinements is fast and easy. Down times are therefore sub-stantially reduced. Furthermore, FRP jackets are thin and require less adaptation of ad-jacent building components.

55

Strengthening method

Incr

ease

in

wei

ght

Dim

ensi

on re

-du

ctio

n

Rad

ial s

tres-

ses

Incr

ease

in

flexu

ral

stre

ngth

Appl

icat

ion

Cor

rosi

on

Steel jackets - - - - -- -- Concrete casing - - - -- -- 0 FRP confinement + + ++* + + ++ Table 3: Comparison of reinforcing methods

* Prestressing without problems 11.2 Reinforced concrete frames For a correct design of seismically endangered structures the structural engineer must be aware of the typical types of damage caused by earthquakes. The most frequent fail-ure modes of reinforced concrete frames are described below. "Short Column" Shear cracking in the concrete occurring at an early stage can be caused by an insufficient transverse load bearing capacity in the beam/column connec-tions, or the beam/slab connections respectively. This crack formation causes the internal stirrup rein-forcement to open and thus leads to failure of the element. The application of FRP jackets to improve the transverse load bearing capacity on the potential plastic hinge regions of the column can prevent this failure mode. Since in addition the overall ductility of the structure must be taken into account, the woven S&P G-Sheet 90/10 or the prestressed S&P A-Strap are specially suited for this type of application. FRP confinement with high modulus C fibres is less suited for this application as the requirements regarding the overall ductility of the structure would be fulfilled to a lesser degree.

Figure 42: „Short Column“

56

Weak junction points

Figure 43: Weak junction point

Junction points that are too weak or those with insuffi-cient load capacity due to reduced cross sections cause further failure modes. Insufficient lap splice lengths of the longitudinal reinforcement at the ends of the columns lead to an additional frequent failure mode. Strengthen-ing can be achieved by CFK laminates applied into longi-tudinal slots or to the surface, with additional wrapping of S&P G-Sheet 90/10 or the prestressed S&P A-Strap over the junction points.

Plastic hinge regions

Figure 44: Shear failure in a plastic hinge region

Due to insufficient shear reinforcement bending beams can exhibit failure in the plastic hinge regions, as shown in Fig-ure 44. Retrofitting is carried out using the high modulus S&P C-Sheet 640 as confinement reinforcement. A further failure mode is the insufficient flexural strength of the beam at mid-span, or near the supports respectively. In this case S&P Laminates CFK are applied into slots or to the surface.

Methodology of retrofitting with FRP In numerous research studies it has been verified that retrofitting of reinforced concrete in the plastic hinge regions provides enhanced load capacity and thus improves the duc-tility of the reinforced concrete frame. The effectiveness of the S&P G-Sheet or S&P A-Sheet in enhancing the ductility has been verified in push-pull tests. The tests further show that GFRP or AFRP provides a higher increase in ductility than a C fibre jacket. This is due to the higher ultimate strain of the glass fibre. Ideally suited for seismic retro-fitting of columns are systems that cause either an increase in hoop stresses on the plastic hinge regions or over the entire length of the column.

57

Figure 45: Retrofitting of a reinforced concrete frame

Large-scale tests indicate that G- or AFRP confinements provide better technical bene-fits and are more economical than steel jackets. In the case of FRP confinement of the entire column or on its ends, concrete failure occurs at larger strains. The reduction of transverse strains provided by the FRP confinement also serves as reinforcement against buckling of the longitudinal reinforcement. Prior to the confinement with FRP, structural repair of cracks in the substrate should be carried out using epoxy resin injection. 12. Explosion and impact protection using S&P FRP systems 12.1 Explosion protection Damage to structures incurred during wars or caused by explosions is frequent. Explo-sion protection is also a requirement of the chemical industry. While explosions in indus-trial buildings can be estimated and the necessary protection thus be designed, estimat-ing the effects of a bomb is impossible. Traditional industrial buildings are often insufficiently reinforced. Masonry structures with little reinforcement are likewise seen in practice. Such structures offer only a minimal resistance to explosion hazards. Conven-tional strengthening methods using steel are costly and their application is time-consuming. FRP provides a time saving and economical solution.

58

Figure 46: Protection of constructional elements sub-ject to explosion hazards

The S&P G-Sheet 50/50 has been specifically developed for this field of application. The sheet with a 50% glass content in the longitudinal and the transverse direction can be ap-plied in several layers. S&P offers also prefabricated GFK laminates for the same purpose. The prefabricated S&P GFK laminates have been tested and approved as protective structural elements by NATO. The GFK lami-nates are structurally glued on to the existing building component using the S&P Resin 220 (Epoxy Resin).

Bi-directionally applied aramid fibres (S&P A-Sheet 120) are likewise suited as explosion protective measure. Due to the favourable mechanical properties of the aramid fibres, and especially their excellent behaviour across the fibre direction, they are ideally suited for this field of application. The high fibre price is often a hindrance to the application of this fibre type. However, AFRP is able to increase the explosion resistance of masonry by a factor of 5 to 10. The related test reports are available from the A fibre manufactur-ers. 12.2 Impact protection The design of columns of for example highway bridges, in practice is often insufficient to sustain the impact of vehicles. The column may not be able to absorb the horizontal loads created by the impact of a truck which could lead to a bridge collapse. Conven-tional strengthening methods, such as concrete casing, are unsuitable because of lack-ing space and for aesthetical reasons. Furthermore, concrete casing requires long traffic down times. An alternative strengthening method is provided by AFRP confinements. The performance of AFRP retrofitted circular columns during the impact of vehicles has been tested by the fibre manufacturers in UK. The AFRP was arranged in orthogonal di-rections, the weight of the sheet layers being 290 gm/m2. Currently, a similar testing pro-gramme is conducted at HTA University of Fribourg (Switzerland) on columns with a square cross section. The tests simulate the impact on a bending beam.

Figure 47: Highway column sustaining a horizontal

impact

The impact of a 30-ton truck at a speed of 75 km/h represents a static load of 1’500 kN that hits the column at approx. 0.75 to 1.5 m above the road surface, plus an addi-tional load of 750 kN hitting the column at approx. 1 to 3 m above surface.

59

The test results show that the energy absorption of columns reinforced in both directions (longitudinal and transverse) is substantially higher than of unwrapped columns. C- and GFRP provided barely any increase in energy absorption. The brittle C and G fibres ex-hibited premature failure due to their insufficient transverse load bearing capacity. The tough aramid fibre, however, was able to yield to the high deformation of the column due to its inherent transverse load bearing capacity. In various test series conducted in UK the behaviour of three AFRP wrapped circular columns has been investigated. The results can be summarised as follows: Column Number of

sheet layers longitudinal / transverse

Max. load [kN]

Max. deflection[mm]

Failure mode

C1 0 0 233 34 Plastic forces in the internal rein-forcement and subsequent compres-

sive strain of the concrete C3 2 2 580 50 Failure of the longitudinal fibres C4 3 2 785 69 Failure of the longitudinal fibres

Table 4: Test results obtained by DUPONT UK (manufacturer of aramid fibres)

A graphical description of the results is shown in Figure 48.

Figure 48: Load-deflection diagram

C1: Reference column C3: 2 longitudinal layers + 2 transverse layers C4: 3 longitudinal layers + 2 transverse layers

In the test the longitudinal aramid fibres were utilised up to failure. Due to the confine-ment in the transverse direction premature buckling of the longitudinal fibres was avoided. Conclusions: • The flexural capacity of a column with longitudinal and hoop directional AFRP wrap

can be substantially improved. • The energy absorption is made possible by the high flexural capacity. • Traditional design concepts can be utilised. • Aramid fibres are suited for retrofitting against collision impacts caused by vehicles.

60

13. Seismic retrofitting of masonry using S&P FRP systems Masonry is an economical construction material that is used on a world-wide basis. Due to its favourable building physical properties masonry will be widely utilised also in fu-ture. Due to varying stone and mortar properties, quality and strength of brickwork are subject to considerable fluctuation. The role of seismic upgrading of masonry in building maintenance is gaining increasing importance. In the past various methods of masonry strengthening have been developed: • Retrofitting using reinforced sprayed concrete and sprayed mortar layers • External prestressing to increase the load bearing capacity • Application of steel reinforcements (bracing of the reinforced concrete frame) The drawbacks of the traditional methods are as follows: • The increase in overall weight of the building will lead to a proportional increase in the

relevant earthquake forces. • As a result of bond problems in the contact region of masonry and sprayed mortar

the reinforcing effect is partially reduced. • The interior space will be reduced. • External prestressing systems cause additional compressive loads. This will often

lead to an overloading of the masonry in the lower floors of a multi-storey building. • Reinforcing systems using steel mats and steel frames are subject to corrosion and

require adequate protection. • Stiffening with steel frames, while providing a substantial increase in resistance, is

problematical from an esthetical point of view and requires extensive application lo-gistics.

With the utilisation of FRP, all the above listed drawbacks are largely avoided. FRP re-pairs require less space and do not corrode. 13.1 Masonry strengthening using the S&P G-Sheet AR 50/50 Strengthening of masonry with glass sheets is adequate in cases where the primary re-quirement is the increase in ductility of a load bearing wall and where the enhancement of the bearing capacity is less important. Seismic retrofitting of load bearing walls should aim at an even distribution of the cracks over the entire wall surface. Enhancement of the ductility means that the cracks should be able to open already under minimal seis-mic action. The S&P G-Sheet 50/50 made of low modulus glass fibres is suited for this field of application. The thus reinforced masonry wall is not subject to additional forces. Strengthening with the S&P G-Sheet 50/50 is carried out on one side of the masonry only (inside or outside). Due to the water vapour permeable S&P Resicem matrix the coating can be applied to the entire surface. The water vapour diffusion coefficient of S&P Resicem is µH20 = 3’000 – 5’000.

61

13.2 Masonry strengthening using S&P Laminates CFK S&P Laminates CFK are able to be bonded diagonally to an existing masonry wall and anchored in the adjacent concrete elements. Tests conducted at EMPA Switzerland in-dicate that the tensile strength of a slot-applied CFK laminate, width 50 mm and thick-ness 1.2 mm, can be anchored at a depth of 25-30 cm into an adjacent concrete ele-ment. Masonry that has been thus reinforced exhibits an elastic behaviour of up to approx. 2/3 of the maximum shear force VA,max (see also Figure 49). As a result of the debonding of the CFK laminate caused by the load transfer from the masonry, the de-formation in the upper strain levels can be heavily increased without causing any sub-stantial increase in the bearing capacity. The brick walls possess large deformation re-serves that contribute to a high ductility. The earthquake resistance of the CFK laminate strengthened wall (sample BW6) could be raised by a factor of 4.3 compared to an unre-inforced wall (reference sample BW5). As shown in Figure 49, the ductility of the wall could be raised by a factor of over 3, with a proportional increase in the bearing capacity by a factor of 1.4. The CFK laminate has been applied to one side of the wall only.

Figure 49: Comparison of FRP reinforced ma-

sonry with reference sample

Figure 50: Load bearing wall with and without openings

13.3 Load bearing wall systems The horizontal earthquake resultant Qacc which is of triangular distribution, the self weight and the live loads on the floor levels, all have to be diverted through the load bearing walls (Figure 50). The earthquake resultant Qacc causes high shear forces in the walls of the lower floors, in combination with low vertical loads. As a result of this unfa-vourable relation between vertical load and shear force, the load capacity is often ex-ceeded. The flexural resistance, by contrast, is usually sufficient. Therefore, the lower floors require a reinforcement that diverts the high shear forces through diagonals sub-ject to tensile and compressive loads. Detailing of the reinforcement mainly depends on the force combination of Mz, Nx und Vy. The requirements for the critical load bearing walls in the lower floors differ from those in the upper floors. Figure 50 shows how the resultants are led around openings. The necessary reinforcement can be designed as a concentrated S&P Laminate CFK or a distributed S&P G-Sheet. Special attention must be paid to the anchoring of the lami-nates and the areas of the load bearing walls that are subject to maximum compressive loads. The example clearly shows that this method of defining stress fields is suited for universal use.

62

63

64

65

66

67

68

16. Literature / Product names • Versuche zur Bestimmung der Werkstoffeigenschaften, der Verbundtragfähigkeit, des Entkop-

pelungsverhaltens von CFK-Lamellen sowie Biegeschubversuche an mit CFK-Lamellen ver-stärkten Platten, Untersuchungsbericht Nr. 8524/5247 des Institutes für Baustoffe, Massivbau und Brandschutz der TU Braunschweig vom20.05.1998.

• Kaiser, H.: Bewehren von Stahlbeton mit kohlenstofffaserverstärkten Epoxidharzen. Dissertation Nr. 8918, ETH Zürich, 1998. • Terrasi, G.P.: Aluminium/CFK Hybride unter thermischer Beanspruchung. Diplomarbeit an der ETH Zürich, Abt. für Werkstoffe. • Vielhaber, Joh.: Versuchsreihe Verstärkung der Querkrafttragfähigkeit von Stahlbetonkonstruk-

tionen, Fachhochschule Potsdam. • Holzenkämpfer, P.: Ingenieurmodell des Verbunds geklebter Bewehrung für Betonbauteile.

Dissertation TU Braunschweig, 1994. • Pichler, D.: Die Wirkung von Anpressdrücken auf die Verankerung von Klebelamellen. Dissertation, Universität Innsbruck Institut für Betonbau, 1993. • Rostasy, F.S., Holzenkämpfer, P., Hankers, Ch.: Geklebte Bewehrung für die Verstärkung von

Betonbauteilen. Betonkalender 1996, Teil II, S. 547 – 638, Berlin: W. Ernst & Sohn, 1996. • Verbundversuche an Doppellaschenkörpern mit CFK-Lamellen und Biegeversuche an mit

CFK-Lamellen verstärkten Platten, Untersuchungsbericht Nr. 8511/8511 – Neu- des Institutes für Baustoffe, Massivbau und Brandschutz Braunschweig vom 05.11.1996.

• Fracture mechanics of Concrete Structures, From Theory to applications, Report of the Tech-nical Committee 90-FMA Fracture Mechanics to Concrete – Applications, Capman and Hall, London, New York, 1989, 188 – 190.

• Rostasy, F.S., Ranisch, E.-H.: Sanierung von Betontragwerken durch Ankleben von Faserver-bundwerkstoffen. Forschungsbericht, Institut für Baustoffe, Massivbau und Brandschutz der Technischen Universität Braunschweig, Dezember 1994.

• Biege- und Schubtragverhalten eines mit geklebten CFK-Lamellen und Stahllaschenbügeln verstärkten Stahlbetonträgers, Untersuchungsbericht Nr.8516/8516 Neu- des Instituts für Bau-stoffe, Massivbau und Brandschutz der Technischen Universität Braunschweig vom 13.05.1996.

• Zilch Konrad, Blaschko Michael, TU München: Verstärkung mit eingeschlitzten CFK Lamellen • Taerwe L., Matthys S., Universität Gent: Inrijgen van Betonkolomnen met vezelcomposietlami-

naten. • Allgemeine bauaufsichtliche Zulassung Deutschland-236.12-54 für S&P Lamellen CFK • Deutsches Institut für Bautechnik: Gutachten Nr. 98/0322 Ingenieurbüro

Prof.Dr.Ing.Dr.Ing.E.h.F.S. Rostasy, Braunschweig • Bemessungsdiskette BOW Ingenieure, Braunschweig • Procédé “Carbone CFK”, Dossier SOTEX No. Ex 1443 • T. Pauly and M.J.N. Priestley: Seismic design of reinforced concrete and masonry buildings • M.J.N. Priestley, F. Seible and G.M. Calvi: Seismic design and retrofit of bridges • T. Ripper, J. Scherer: "AVALIAÇÃO do DESEMPENHO de PLÁSTICOS ARMADOS com FOL-

HAS UNIDIRECCIONAIS de FIBRAS de CARBONO como ELEMENTO de REFORÇO de VIGAS de BETÃO ARMADO", IBRACON 41stCongress, 1999, Salvador, Bahia, Brasil.

• TU Porto (P) Pilares De Betao Armado Reforcados Com Laminados De Fibras De Carbono. • Weidner, J.; Köhler, W.; Krams, J.: Verstärken von Betonbauteilen mit geklebter Bewehrung.

Beton- und Stahlbetonbau 95, 9/2000, S.531 – 536 • Non-Metallic (FRP) Reinforcement for Concrete Structures, Volume 1, Proceedings of the

Third International Symposium on Non-Metallic (FRP) Reinforcement for Concrete Structures, October 1997.

69

• Verstärken von Mauerwerk mit Faserverbundwerkstoffen, Dissertation ETH Nr. 10672, Gregor Schwegler – Koller, 1994.

• Seismic Design and Retrofit of Bridges, M. J. N. Priestley, F. Seible and G. M. Calvi, 1996. • Seismic Design of Reinforced Concrete and Masonry Buildings, T. Pauly and M. J.N. Priestly,

1991. • Design for Earthquakes, James Ambrose and Dimitry Vergun, 1999. • Torsion und Duktilitätbedarfs bei Hochbauten unter Erdbebeneinwirkungen, Dissertation ETH

Nr. 13738, Aloïs Sommer, 13.09.2000. • Fastenings for Seismic Retrofitting, State of the Art Report, Comite Euro-International du Beto

(CEB), 1997. • Evaluation of Seismic Retrofit Methods for Reinforced Concrete Bridge Columns, Terry J, Wipf,

F. Wayne Klaiber and Frank M. Russo, Iowa State University, Department of Civil and Con-struction Engineering, 28.12.1997.

• Erdbebensicherung bestehender Bauwerke und aktuelle Fragen der Baudynamik, Referate der D-A-CH-Tagung 1997, Dokumentation D 0145, 25./26. September 1997.

• Kevlar Aramid Fiber for External Strengthening & Repair of Concrete Structures, René Pinzelli, Structurl Faults & Repair 1999.

• Fibre Reinforced Plastic Strengthening of Bridge Suppprts to Resist Vehicle Impact, J.R. Cuninghame and B. Sadka, Proceedings of the 20th International SAMPE Europe Confer-ence of the Society for the Advancement of Material and Process Engineering, Paris, April 1999.

• FRP Design Guide, S&P Clever Reinforcement Company, Brunnen, Switzerland, June 2000. • M.J.N. Priestley, F. Seible and E. Fyfe: - Column Seismic Retrofit Using Fiberglass/Epoxy

Jackets – Proceedings 1st International Conference on Advanced Composite Materials in Bridges and Structures, 1992, pp 287-298

• Wang Yung-Chih – Retrofit of Reinforced Concrete Members Using Advanced Composite Ma-terials – Research Report 2000-3, Department of Civil Engineering, University of Canterbury, February 2000, ISSN 0110-3326

• Dodd, L.L. and Restrepo-Posada, J.I. – Model for predicting cyclic behaviour of reinforcing steel – Journal of Structural Engineering, ASCE, vol 121, No. 3, pp 443-445, 1995

• Mander, J.B., Priestley, M.J.N. and Park, R. – Seismic Design of Bridge Piers - Research Re-port 84-2, Department of Civil Engineering, University of Canterbury, New Zealand, 442 pp,1984

• Sheikh, S.A. and Uzumeri, S.M. – Strength and ductility of tied concrete columns – Journal of Structural Division, ASCE, ST5, pp. 1079-1102, 1982

• Mander, J.B., Priestley, M.J.N. and Park, R. – Theoretical stress-strain model for confined con-crete – Journal of Structural Division, ASCE, vol 107, No. ST11, pp 2227-2244, 1998

• Park, R. and Pauley, T. – Reinforced Concrete Structures – John Wiley ans Sons, New York, 769pp

• American Concrete Institute – Building Code Requirements for Reinforced Concrete (ACI 318-95) and Commentary (ACI 318R-95) – Detroit, Michigan, 1995

• The Concrete Society (UK): Technical Report No. 55 – Design guidance for strengthening con-crete structures using fibre composite materials – ISBN 0 946691 843, 2000