SP 209-17

Self-Consolidating Concrete Piles Confined in FRP Tubes

by H. El-Chabib, M. Nehdi, and M. H. El Naggar

Cast-in-place deep foundations such as drilled shafts and piers are often subjected to two sources of problems. First, the integrity and uniformity of the cross-sectional area of these structural elements cannot be assured using n o d concrete because of limited accessibility and visibility during construction. Can- ties and soil encroachments leading to soil pockets can jeopardize their load-bear- ing capacity. Second, corrosion problems of steel reinforcement in deep founda- tions have heen costly, requiring annual repair costs of more than $2 billion in the US alone. To address these two challenges, a novel technology for the construction of drilled shaft concrete piles is proposed in this study. Self-oonsoiidating con- crete, a material that compacts under its self-weight without vibration and without bleeding or segregation. is used to assure the structural integrity and uniformity of the cross-sectional area of deep foundations. The self-consolidating concrete is cast into FRP envelopes. which provide corrosion-resistant reinforcement. This pa- per presents results of a laboratoiy investigation on the mechanical performance of these novel piles including the effect of using expansive cement and shrinkage-re- ducing admixtures to enhance the FRP tube-concrete interfacial bond.

Keywords: bending; compression; concrete; fiber-reinforced; piles; plastic; self-consolidating; stress-strain

Hassan El-Chabib is a graduate research assistaot in the Department of Civil & Environmental Engineering, University of Western Ontario (UWO), London, Ontario, Cana&. His research interests include concrete technology and attificial neural networks applications in civil engineering.

ACI member Moncef Nehdi is an assistant professor in the Department of Civil & Environmenial Engineeing at UWO. He is a member of ACI Committees 225, Hydraulic Cements; 236, Material Science and E803, Faculty Network. His research interests include modeling the behavior of cement-based materials, recycling by-products and durability and repair.

M. Hesham El Naggar is an associate professor in the Department of Civil & Environmental Engineering at UWO. His research interests include analysis and design of deep foundations, machine foundations, earthquake- resistant structures, and use of special concretes in geotechnical applications.

INTRODUCTiON

The ultimate load-bearing capacity of cast-in-place concrete piles is significantly affected by the compressive strength of concrete, the pile cross-section, the amount of steel reinforcement and various soil properties. Both the compressive strength of concrete and the integrity of the cross- section depend on the consolidation of fresh concrete during placement.

Generally, concrete is consolidated with mechanical vibrators, a process that is operator-sensitive. in deep concrete piies where accessibility and visibility are limited, consolidation is particularly difficult. Over- consolidated concrete might segregate or destabilize soil panicles around the shaft wall causing it to collapse inside the concrete core. Under- consolidated concrete can have voids and cavities, leading to air and soil pockets and exposure of steel reinforcement to corrosion attack. Consequently, the integrity and uniformity of the pile’s cross-section often cannot be assured, concrete strength is adversely affected and the load- bearing capacity of the pile is jeopardized.

To alleviate both the problem of air cavities and soil pockets in the concrete and corrosion of steel reinforcement, a novel technology for the construction of drilled-shafi concrete piles is proposed in this study. Self-

consolidating concrete (SCC), a material that consolidates under its own weight with little or no vibration and without exhibiting segregation or bleeding, is used instead of normal concrete to assure the structural integrity and uniformity of the cross-sectional area of deep foundations. Self- consolidating concrete is cast into a fiher-reinforced plastic (FRP) envelope, which acts as stay-in-place structural formwork and provides a protective jacket for concrete and steel reinforcement.

In addition, the FRF’ envelope acts as a protective jacket in aggressive corrosive environments such as the case of marine piles ( I ) and piles constructed in soils rich in chloride ions and sulfate salts (2), and may provide total or partial replacement of longitudinal and shear steel reinforcement. More importantly, the FRP tube provides a lateral confinement to the concrete core in piles and significantly enhances the compressive strength of concrete and its ductility (3), (4).

The confinement of concrete is the restraint of lateral deformation of concrete under axial loading. Extensive research bas been conducted on concrete-filled steel tubes for use as piles and columns (5), (6). However, FRP tubes provide several advantages over steel tubes when used as c o n f i g protective jackets for concrete piles. in addition to their high- strength to weight ratio, FRP materials are known for their high corrosion- resistance. Unlike the confining effect provided by steel tubes, which reaches a threshold value once the steel yields, FRP tubes provide a continuously increasing confining pressure until failure, which adds to both the ultimate compressive strength and ductility of the concrete member (7).

The focus of this study is to investigate the confinement effect of glass fiber-reinforced plastic (GFRP) tubes on the stress-sirain behavior of short self-consolidating concrete cylindrical piles under uniaxial compression and iransverse loading. Specimens made both of normal concrete (NC) and SCC were tested. Moreover, because the interfacial contact between the concrete core and the confining tube is extremely important to achieve a composite behavior, especially under transverse loading, SCC and NC specimens prepared using Type 10 Canadian ordinary portland cement (ASTM Type I) and expansive cement (Type K) along with a shrinkagereducing admixture were investigated. Both ends of specimens tested under transverse loading were monitored for any slippage between the concrete and the confining tube.

EXPERIMENTAL PROCEDURES

A total of 24 cylindrical specimens were prepared and tested in this shidy. Twelve 150 x 300 nun short cylindrical columns were tested under uniaxial compression to investigate the confinement effect of GFRP on axially loaded normal concrete and self-consolidating concrete piles made with either ordinary portland cement (OPC) or expansive cement (EC). Twelve 150 x 1100 mm cylindrical beams were tested under transverse load to investigate the confinement effect of GFRP tubes on normal concrete and SCC piles under transverse loading and the effect of using expansive cement and a shrinkage-reducing admixtwe on the interfacial bond between the concrete and the confining tube.

Short-column specimens included 4 plain concrete control cylinderr (two NC and two SCC) and eight concrete cylinders confined by GFRP tubes. The confined specimens included 2 cylinders made of normal concreie using OPC, 2 cylinders made of SCC using OPC and two cylinders of each (NC and SCC) made using expansive cement. Similarly, the beam specimens included 4 plain concrete control beams (two NC and two SCC) and 8 concrete beams confined by GFRP tubes (2-OPC-NC, 2-OPC-SCC, 2-expansive-cement-NC and 2-expansive-cement-SCC).

Materials

Concrete specimens were prepared from four different concrete mixtures proportioned to achieve similar 28-day compressive strengths. Proportions of concrete mixtures and their correspodig compressive strengths are shown in Table I . Canadian Type 10 OPC (ASTM Type I) was used to prepare the normal concrete (NC) mixture. The SCC mixture was prepared using the same iype of cement incorporating proportions of fly ash (FA) and grannlated blest furnace slag (GBFS) along with a superplasticizer and a viscosity-modifying admixture (VMA). Two other concrete mixtures (NCE and SCCE) having the same pmportions as the NC and SCC mixtures were prepared using expansive cement (EC) instead of Type 10 OPC in addition to a shrinkage-reducing admixture. Anhydrous tetracalcium hialuminate sulfate based Type K cement was used for the expansive cement along with a commercially available shrinkage-reducing admixture.

FRP tubes consisted of a filament-wound angle .ply with bi-directional glass-fiber at -t 55" winding angle with respect to the longitudinal axis of the tube. They had an inside diameter of 150 mm and a uniform thickness of 6 mm. The physical and mechanical properties of the GFRP tubes are listed in Table 2.

Specimen Preoaration and Testine Procedures

Uniaxial Comuression Test -- plain concrete and concrete-filled F W tube composite cylinders were prepared according to ASTM C 192-90a (Standard Practice for Makiog and Curing Concrete Test Specimens in the Laboratory) but no rodding or vibration was performed on SCC specimens. All concrete mixtures were prepared using the same mixing sequence in a laboratory rotary mixer. Ail specimens were moist-cured in a curing mom with 100% relative humidity for 27 days, then they were all capped with a sulfur solution at both ends and stored at room temperature (about 22°C) until testing at 91 days from casting. Table 3 shows the properties of all cylinders used for the uniaxial compression test.

The average axial and circumferential strains at mid-height on the exterior surface of each specimen were measured by means of strain gauges. The axial deflection (axial shortening of the column) was measured using an LVDT. Each specimen was instrumented with four 10-mm strain gauges located at its mid-height and attached to its exterior surface. Two gauges were attached in the axial direction at 180' apart to measure the average axial strain and the other 2 gauges were attached in the circumferential (hoop) direction at 180' apari and 90" from the axial strain gauges to measure the average radial strain as shown in Fig. 1. The compression test was performed using a 4600 !& MTS machine at a loading rate of 4.5 kN/sec and the axial load was applied to the entire cross-section of the composite cylinder. Fig. 2 shows the typical test set-up of a specimen under uniaxial compression.

Flexural Test -_ all beam specimens were prepared in a similar manner to the short-column specimens. However, they were moist-cured for 27 days in a curing room at 100% relative humidity and tested at 28days after-. casting. Table 4 summarizes the properties of ail beam specimens used for the transverse loading test.

AU beam cylinders were instrumented with six 10-nun strain gauges attached to the exterior surface of either the FRF' tube in the confuied specimens or the concrete surface in the unconfined ones. Four strain gauges were installed at the mid-span of the beam, two at the top section of the specimen to measure the average compression strain, and two at the bottom section to measure the average tensile strain. One strain gauge was attached at a distance di2 from each support at the specimen's mid-height and at 45" from the tube's longitudinal axis to measure the average maximum shear strain at that point, (d = the outside diameter of the specimen). The maximum beam deflection was measured using an LVDT located right under the bottom of the central mid-span point of the beam (Fig. 3).

All beam specimens were subjected to four-point flexural loading. The distance between the two supports (free span) was 1000 mm and the loading points were 250 mm apart in the specimen's m i d - d o n . The applied load was transferred to the specimen through a steel I-section beam resting on two 25-mm diameter steel rollers located at the loading points. Two other 25-mm diameter steel rollers were placed at the supports. Four 10-mm thick, SO-mm wide and 162-mm inside diameter half-steel rings having an outside rectangular shape were manufactured and rubber coated to distribute the load to the surface of the cylindrical tube at the loading points and supports. Fig. 3 shows a schematic representation of the bending test set-up. The transverse load was applied using a 1300 kN hydraulic jack in a displacement control mode at a rate of 6 mdmin for composite beams and 0.5 d m i n for the unconfined concrete beams. Fig. 4 shows the laboratory test set-up of a beam specimen subjected to transverse load.

RESULTS AND DISCUSSION

Uniaxial Cotnureasion Test

The axial load-defomation curves of composite short-column specimens under uniaxial compression can be characterized hy three different regions. Two linear stages (first and third regions) connected by a nonlinear transition stage (second region). The first region is dominated by the behavior of the concrete core and the third one is dominated by the behavior of the FñP tube as shown in Figs. 5 and 6. All conñned specimens behaved similarly during the first and third regions of the loading process. At a b u t 50 % of the ultimate load, cracking noises were heard indicating the

progress of failure of the concrete core. followed by a quite steady period during the transition region. White lines started to develop along the fibers at 55" with respect to the longitudmal axis of the tube at about 60 % of the ultimate load, indicating the development of stress in the fibers as shown in Fig. 7. The white colored area along the fibers increased with increasing load to cover most of the middle third of the specimen followed by fracture of fibers and a sudden failure. For all composite specimens, failure was initiated within the middle third of the specimen height and progressed towards its top and bottom ends. Fig. 8 shows a typical failure mode of a concrete-filled FRP tube under uniaxial compression.

The test results for all short-column specimens are summarized in Table 5 and the axial load- deflection curves for NC and SCC specimens along with a hollow FRP tube are presented in Figs. 5 and 6, respectively. Although the behavior of NC and SCC composite cylinders were very similar, it is clear from both Figs. 5 and 6 that the first region of the composite cylinder axial load-deformation response coincided with that of unconfined concrete up to 30 % and 50 % of the unconhed concrete strength for NC and SCC, respectively. However, in both cases (NC and SCC) the confinement effects on concrete strength and axial deformation were similar and equal to 33% (increase) and 13% (decrease), respectively. The confinement effect was calculated as the difference between the concrete failure in confined and unconfined cylinders. The failure of the concrete core in a confined specimen was defined as the point at which the axial load-deformation curve of the composite FRP-concrete specimen shifted from a linear (stage I) to a non-linear (stage 2) behavior (Figs. 5 and 6). The ultimate axial-load and axial-deformation of composite cylinders in both cases (NC and SCC) were found to be about 2.5 and 12 times higher than those of the unconfined cylinders, respectively.

The only significant difference between the confined NC and SCC behavior was in the transition region. Transition between the first and third regions in the response of NC was progressive, while for SCC the transition was more sudden as shown in Fig. 9. This indicates that for SCC there was a short time lag between the instant when the concrete core failed and when the effect of the FRP was mobilized. This suggests that higher shrinkage may have occurred in the case of SCC and made the FRP-concrete interface less tight. This is confirmed by the fact that SCC made with expansive cement and a shrinkage-reducing admixture did not show this type of transitional phase in its axial loaddeformation curve (Fig. 6).

The stress-strain response curves for plain NC and SCC specimens and concrete-filled FRP composite specimens are plotted in Fig. 10, in which the fust region of the axial stress-strain response of the composite qylinders coincides with the unconíiied concrete response in both cases (NC and SCC-filled FRP tubes). For the same axial deformation value, the higher axial stress obtained in the case of SCC-FRP composite is attributed to the fact that the 91-day compressive strength of plain SCC is slightly higher than that of plain NC. For the axial stress-circumferential strain curves, Fig. 10 shows that the stiffness of the unconfined cylinders is slightly higher than that of the confined ones in both NC and SCC members. This behavior is likely due to the fact that the axial load was applied to the entire cross- section of the composite cylinder, and because concrete is stiffer than FRP, a higher value of circumferential strain is introduced at the outside surface of the FRP tube under the same axial load. The effect of using expansive cement instead of OPC in NC and SCC is demonstrated in Figs. 11 and 12 in which no significant change was observed in the coníiiement effect on concrete strength and ductility under uniaxial compression. Again the higher stresses in the response cuwes of NCE (normal concrete with expansive cement) are attributed to the fact that the unconfned compressive strength of NCE was slightly larger than that of the uncodmed compressive strength of NC.

Flexural Test

The load-deflection curves of concrete-filled F W composite tubes subjected to transverse load can also be characterized by three different regions, as it is the case for concrete-filled FW tubes under uniaxial compression. However, the strength and stiffness of the FRP material dominated the behavior of such beam specimens during the loading process in all three stages. Since no steel was used to reinforce the concrete core in the unconfined beam specimens. their ability to cany transverse load was almost negligible. However, in the case of concrete-filled FRP composite beams, the concrete core contributed substantially to the overall strength and ductility of the composite specimens. Figs. 12 and 13 illustrate the load- deflection responses of confined and unconfied NC and SCC specimens along with the response of a hollow FRP tube, ail subjected to transversal load. It is shown that the capacity of the FRP tube in sustaining transverse load increased by 175 % from 45 kN (ultimate load of hollow tube) to 124 kN (ultimate load of concrete-FRP composite member) and the maximum deflection of the tube at mid-span was also increased by 133 % from 21 mm in the case of a hollow tube to 49 mm in the case of an FRP-concrete composite beam. SCC specimens behaved relatively in a similar manner to

NC specimens and no major differences were observed during the loading process. Fig. 13 provides a comparison between the behavior of NC and SCC specimens under transverse load.

Slippage between the concrete core and the FRP tube may compromise the ultimate load capacity of the composite member, especially under transverse load. Therefore, an attempt was made to strengthen the interfacial bond between the two materials using expansive cement instead of OPC with the addition of a shrinkage-reducing admixture. This type of concrete will expand rather than shrink and therefore can create an active hoop pressure against the intemai wail of the FRP tubes, therefore developing a better bond between the concrete core and the confining tube. Since such composite tubes are usually intended for use underground (saturated soil), subsequent shrinkage is considered less significant. The shrinkage-reducing admixture will reduce shrinkage strains, further enhancing this bond. The behavior of specimens prepared with such type of concrete is described in Figs. 13. It is shown that in the first and second stages of the response curve and at the same deformation, the load capacity of a GFRP tube filled with concrete using expansive cement is noticeably higher than that of a tube filled with concrete made of OPC, indicating that the specimen made with expansive cement was stiffer. However, the ultimate load and deflection of such specimens at failure was slightly lower than those of specimens made using OPC. This is believed to be due to the fact that the fibers in the FRP- concrete composite tube made using expansive cement were already prestressed due to the expansion of concrete before the loading process Started.

Another attempt was made to prevent slippage by anchoring the FRP tube to concrete at 200 mm from each of its ends by inserting a 12-mm diameter steel bar in each of two boles drilled through the FRP before casting the concrete. It was later observed during testing that a local failure occurred in the concrete around the steel bars due to stress concentrations, and no major contribution to prevent slippage was noticed (Figs. 12). In fact, for all tested specimens slippage between the concrete and the FRF' at both ends of each beam varied only between 1 mm and 3 mm. During the loading process, it was observed that each time slippage occurred between the concrete and the FRP, the load dropped, which explains the existence of harmonic events in the loaddeflection curves of Figs. 12 and 13. All concrete-filled GFRP tubes shared the same failure mode. White lines along the fibers siarted to form at mid-span of the bottom section of the FRP tube and progressed towards the ends. Tensile cracks started to appear on the bottom section of the tube and under both loading points and progressed towards the upper

section until a major crack was developed to cause a sudden failure as shown in Fig. 14.

CONCLUSIONS

This study investigated the behavior of concrete-filled GFRP tubes under both uniaxial compression and transverse loading for possible use in deep foundations (drilled-shaft piles). Special focus was on: i) the axial load- deformation response and axial stress-strain response of GFRP tubes filled with either NC or SCC under uniaxial compression; ii) the load-deformation behavior of such composite tubes under transverse load; and iii) the effect of using expansive cement and anti-shrinkage admixtures in concrete on the interfacial bond between the FRP tube and the concrete core. The following conclusions em be drawn from this investigation.

I . SCC-filled GFRP tubes had a similar behavior to that of NC filled- GFRP tubes under both uniaxial compression and transverse load.

2. The only significant difference between the behavior of NC and SCC-filled GFRP specimens was in the transition region of the response curves, in which the shift h m a linear to a non-linear behavior in the load-deformation and stress-strain curves subsequent to the failure of the concrete core was more sudden for SCC-filled FRP specimens. This is believed to be due to autogenous and self- dessication shrinkage, which is higher in SCC due to a large volume of water and cement paste and lower volume of coarse aggregates. However, the use of expansive cement and shrinkage-reducing admixtures made the behavior of SCC-filled FRP tubes similar to that of NC-filled FRP tubes.

3. The use of expansive cement in concrete delayed the occurrence of slippage between the FRP tube and the concrete core, creating a somewhat better bond between the two materials, but did not fully prevent it. Likewise, the use of localized lateral steel bars placed through the FRP tube and concrete core did not prevent slippage. Shear connectors or ribs placed inside the FRP tubes may provide better performance (8).

4. FRP tube confinement of concrete cyluiders increased their ultimate load by 2.5 times and their axial deformation by 12 times under uniaxial compression. It also enhanced their ultimate load by 20 times and their mid-span deflection by 100 times under transverse load.

REFERENCES

I . Sen, R., Shahawy, M., Rosas, J. and Sukumar, S., “Durability of AFRP and CFRP Pretensioned Piles in Marine Environment,” Non-Metallic (FRP) Reinfoment for Concrete Structures, Proceedings of the Third International Symposium, Vol. 2, 1997, pp. 123-130.

2. El-Nawawy, O. and Kandeil, A., “A New Reinforcing Material for Concrete in [he Arabian Gulf;” Proceedings of the 2“ International Conference on Deterioration and Repair of Reinforced Concrete in the Arabian Gulf, CIRIA, Bahrain, 1987, pp. 185.

3. Mimiran, A. and Shahawy, M., “Rehavior o/ Concrete Columns Confined by Fiber Composites,” Journal of Structural Engineering, Vol. 123, No. 5, 1997, pp. 583-589.

4. Samaan, M. S., “An Analytical and Experimental Investigarion of Concrete-Filled Fiber Reinflrced Plastics (FRP) Tubes.” Ph.D. Thesis, University of Central Florida, Orlando, FL, 1997,220 p.

5. Knowls, R. B., and Park, R., “Strengih of Concrete Filled Steel Tubular Columm,” ASCE, Journal of the Structural Division, Vol. 95, ST12, 1969, pp, 2565-2587.

6. Mander, J. B., Park, R,, and Priestly, M. J. N., “Theoretical Stress-Strain Model for Confined Concrete,” Journal of Structural Engineering, ASCE, l l4(8), 1988,pp. 1804-1826.

7. Samaaq M., Mimiran, A. and Shahawy, M., “Model of Concrete Confined by Fiber Composites,” Journal of Stnictural Engineering, Vol. 124,NO. 9, 1998,pp. 1025-1031.

8. Mimiran. A., Shahawy, M., Samaan, M., El Echary, H., Masirapa, J. C., Pico, O., “Effect of Colomun Parameters on FRP-Confined Concrete, ‘I Journal of Composites for Construction, Vol. 2, No. 4, 1998, pp. 175-185.

NOTATiONS

a V = Poisson ratio. &,

4,

41, A-

&o

AS?íA = Anti-shrinkage admixtures. C =Cement. D E EC =Expansive cement. FA =Fly ash. G =Gravel. GBFS =Granulated blast furnace slag. GFRP =Glass fiber-reinforced plasiic. L = length of specimen. NC = Normai concrete. NCE

OPC =Ordinary portland cement. P, P , Put, Pm S =Sand. SCC = Self-compacting concrete. SCCE = Self-compacting concrete with expansive cement and anti-

SP = Superpiasticizer. VU4 = Viscosity-modifying admixtures. w =water. W/C =water to cement ratio. f; f e u t

=Winding angle of glass fibers in the FRP tubes.

=Maximum axial deflection of unconfined cylinders at concrete

=Maximum axial deflection of confined cylinders st concrete

=Ultimate axial deflection of confined cylinders at failure. =Maximum mid-span deflection of specimen subjected to pure

=Maximum axial strain of unconfined cylinders. =Ultimate axial strain of confuied cylinders.

failure.

failure.

bending.

=Inside diameter of FRP tubes. = Modulus of elasticity of FRF' tubes.

=Normal concrete with expansive cement and anti-shrinkage admixtures.

=Maximum axial load of unconfined cylinder at concrete failure. =Maximum axial load of confined cylinder at concrete failure. =Ultimate axial load of codmed cylinder at failure. =Maximum transversal load of cylinder subjected to pure bending.

shrinkage admixtures

=Compressive strength of unconfined concrete specimen =Ultimate compressive strength of confined concrete specimen =Thickness of the FRF' tube.

Table I - Mixture proportions and compressive strengths of cuncrete batches W/B' Mixtures Component (kgím')

NC"

SCC:

NCE'

SCCE'

(mm)

15.0

ímm) Strength E * Strength E* Lateral/ Transverse/ (MPa) (GPa) (MPa) (GPa) transverse lateral

6.0 155' 53.5 60.0 8.5 193.0 10.5 0.39 0.5

cr _____ I050

830

1050

850 2 1 0.04

4

4

___ 28-d

34.5

35.5

37.0

38.5

-

-

_____ 90-d

39.5

43.8

42.0

45.4

__

',* See NOTAI'IONS

Table 2- Physical and mechanical properties of FRP tubes

D* I t* 1 a* I Glass content (%) I Axial direction 1 Hoop direction 1 Poisson's ratio (v)

* See NOTATIONS

SCCE SCCE

FRP Hollow

Cylinder's I Concrete 1 f, (MPa) No. of conf. No. of unconf. I' ~

90-d cylinders cylinders (mm) 39.5 2 2 6

43.8 2 2 6

42.0 2 6

45.4 2 6

2 6

D' L'

(mm) (mm) 150 300

150 300

IS0 300

IS0 300

150 300

Table 4- Propenies of cylindrical beam specimens prepared for flexural test

2

2

2

Cylinder's I Concrete 1 f , (MPa)

2 BSCC

BNCE NCE

F W Hollow 1 -

__ 90-d 39.5

43.8

42.0

45.4

-

___

l l00

~

I see NOTATION.

Table 5- Test results for specimens subjected to uniaxial cc

6.1

45.0

SCC

NCE SCCE

FRP

*See NOTATIONS

0.45

21.0

7

P.,,

WN) ____ 1770

1736

I743

1784

420 ____

xession

0.05

0.06

Table 6- Test results for specimens subjected to transverse load

NCE 37.0

SCCE 38.5

FRP

Unconfined specimens 1 Confmed specimens

121.5

118.5

118.0

-ímm) 49.7

50.0

39.0

53.5

2 axial strain gauges

2 circumferential strain gauges (1 80' apart)

-.

Fig. 1- Illustration of specimen used in uniaxial compression

Fig. 2- Test set-up for uniaxial compression

Cancun Conference Proceedings 313

I-section steel beam kN hydraulic

Strain gauges BEAM SPECIMEN

T

Free span = 1 000 mm I 1 1 I I Total lengih = 1100 mm I

Fig. 3- Illustration of test specimen subjected to transverse load.

Fig. 4- Test set-up for flexural test.

o 2 4 e B 1 0 12 14 16 18 20

Axhl dellmuon (mm)

Fig. 5- Axial loaddeformation respoiu of NC cylinders and FRF' tube

- 1 I

Ix!Ad.f*cuon (Mn)

Fig. 6- Axial load-deformation response of SCC cylinders and FRP tube

Fig. 7- Highly stressed area (white lines) along fibers.

Fig. 8- Typical failure of a concrete-filled FRP tube under uniaxial compression.

Fig. 9- Axial load-deformation response of NC and SCC cylinders

25

Fig. 10- Stress-strain response of NC and SCC cylinders under axial load, (strain gauges failed at their maximum limit of 0.02).

O

02

OP

o9 0 P -

os r ooi

O Z I

001

140

120

1W

a -I 60

40

20

o O 10 20 30 40 50 60

C."t.rd.RKlion (mm)

Fig. 13- Load-deflection response ofNC and SCC specimens under transverse load.

Fig. 14- Typical failure mode of concrete-filled GFRP tube under transverse load.