1PCI Journal | January–February 2008

Effects of freezing and thawing cycles and sustained loading on compressive strength of precast concrete composite piles Amir Fam, Andrew Kong, and Mark F. Green

The concept of precast-concrete-filled, fiber-reinforced-polymer tubes (CFFTs) has been introduced as a viable alternative to con-ventional reinforced concrete members, particularly in structural applications in harsh marine environments. Figure 1 shows a se-verely deteriorated conventional concrete pile in such an environ-ment. The prefabricated composite tube in the CFFT system resists corrosion and protects the concrete core from various deteriora-tion and damage mechanisms while eliminating the need for steel reinforcement of the concrete and providing excellent confinement to the concrete core.1

This paper reports the findings of an experimental investiga-tion into the structural response of CFFTs subjected to combined sustained axial loads and freezing and thawing cycles before being tested to failure under axial compression. CFFT control specimens subjected to either sustained load only or freezing and thawing cycles only were also tested, and the results were compared.

Background information

Numerous field installations using CFFT piles have been suc-cessfully completed along the east and west coasts of the United States.2 In these applications, the piles were typically driven into the ocean floor and extended above the water surface to serve

Editor’s quick points

n  As part of this issue’s theme on service life and dura-bility, this paper demonstrates that commercially avail-able glass-fiber-reinforced-polymer tubes can serve as protective jackets for concrete in severe environments.

n  This paper serves to reassure designers about the long-term durability of precast concrete composite piles subject to freezing and thawing and sustained load.

January–February 2008 | PCI Journal2

a variety of structural purposes: as fender piles, mooring piles, bridge pier protection, and supports for platforms and wharves (Fig. 2). The CFFT system has also been used in bridge piers partially submerged in rivers (Fig. 3).3 Some of these ap-plications have been in the northern part of the country, where freezing is likely during the winter months.

Structural performance of the CFFT system under axial compression, bending, and combined load-ing has been studied extensively at room tempera-ture.4,5 In addition, the long-term performance of CFFT axial members under sustained loads at room temperature, including the creep and shrinkage ef-fects, has also been studied.6

The effect of freezing and thawing cycles has only been studied for concrete cylinders wrapped with fiber-reinforced-polymer (FRP) sheets to simulate reinforced concrete columns retrofitted by FRP wraps.7,8,9 In all of these studies, the freezing and thawing cycling was conducted without sustained loads. Recently, the combined effect of freezing and thawing exposure and sustained axial load-ing was studied by the authors for FRP-wrapped cylinders.10

It is worth noting that FRP wraps are used to retrofit old columns where the concrete has already undergone shrinkage, whereas in CFFTs, shrink-age of the concrete occurs after the tubes have been filled. Also, fibers in FRP wraps are typically circumferentially oriented, whereas in FRP tubes, a portion of the fibers is oriented longitudinally or at an angle, which makes Poisson’s ratio of the FRP tube substantially different from that of FRP wraps. This effect, combined with concrete shrinkage, could affect the interface contact pressure between the concrete and the tube, hence the watertightness of the system.

The effect of long-term submersion in water at room temperature for CFFTs has been studied.11 No studies, however, have been conducted on the freezing and thawing performance of unloaded or axially loaded CFFTs. Freezing and thawing effects are important in cold regions and vital for coastal regions that may experience repeated cycles annu-ally. In addition, CFFTs typically resist sustained axial loads in service, but most durability studies have neglected the effect of load. The synergy between load and freezing and thawing effects may be more severe than either effect on its own. Thus, this study investigates these combined effects.

Figure 1. Pictured is degradation of conventional concrete piles in marine environments. Photo adapted from M. G. Iskander and A. Stachula's 1999 article “FRP Composite Polymer Piling: An Alternative to Timer Pilling for Water-Front Applications” in Geotechnical News.

Figure 2. This is an example of concrete-filled, fiber-reinforced-polymer tube pile in marine applications. Photo adapted from Fam.2

Figure 3. This is an example of concrete-filled, fiber-reinforced-polymer tube pile application in bridge piers. Photo adapted from Fam.3

were kept at room temperature without sustained loads (R), also for the same duration.

In addition, for each concrete type, three standard plain cylinders were subjected to the same freezing and thawing regime as the CFFT specimens (C-F), three specimens were kept at room tem-perature (C-R), and two cylinders were kept at room temperature for the same duration of freezing and thawing cycling but were submerged in water for a period of 10 days (C-RW), which was equivalent to the cumulative thawing time of the specimens sub-jected to freezing and thawing cycling in which water was used for thawing. This practice was adopted to represent any curing effect that the thawing water may have had on strength.

Additional plain concrete control specimens were also used to examine the strength at various time intervals, which was needed as a guide for planning the test program. Standard cylinders were used because the sizes of the cylinders were comparable to those of the tubes.

Materials

Prefabricated glass FRP (GFRP) tubes of 6.57 in. (167 mm) diameter and 0.211 in. (5.35 mm) wall thickness were used in this study. The filament-wound tubes included fibers oriented in the longitudinal and hoop directions at angles of 6.5 and 5 degrees, respectively. The ratio of fibers oriented in the hoop and longi-tudinal directions was 2:1, which resulted in tensile strength and modulus of 31.91 ksi (220 MPa) and 4206 ksi (29 GPa) in the hoop direction, based on manufacturer data.12 The tubes were cut into 13.15-in.-long (334 mm) sections.

Experimental program

The objectives of the study were to examine the individual and combined effects of freezing and thawing cycles and sustained axial compressive stresses on the residual strength of CFFT piles and to compare two different types of concrete—a low-strength (3.2 ksi [22 MPa]) normalweight and a medium-strength (5.9 ksi [41 MPa]) lightweight concrete—under these conditions.

The following sections provide descriptions of the test specimens, materials, fabrication processes, preparations of specimens, procedures for freez-ing and thawing exposure and sustained loading, instrumentation, and post-exposure testing.

Test specimens and parameters

Table 1 provides a summary of the test matrix. The study included both normalweight (nw) and lightweight (lw) concrete for filling the tubes. For each type of concrete, three specimens were simul-taneously subjected to freezing and thawing cycles and sustained axial compression load (FS), two specimens were subjected to the same freezing and thawing cycling regime without sustained loads (F), three specimens were kept at room temperature under sustained loads (RS) for the same duration as the (FS) and (F) specimens, and two specimens

3PCI Journal | January–February 2008

Table 1. Summary of test matrix and test results

Specimen name

Specimen type

Concrete density

Exposure type

Sustained loading

Number of specimens

Unconfined compressive strength, ksi

Ultimate confined strength, ksi

Individual Avg. Individual Avg.

FS-nw

CFFT

nw

FYes 3

n.a.

11.1, 10.0, 10.0 10.4

F-nw No 2 9.0, 8.77 8.8

RS-nwR

Yes 3 9.3, 9.0, 10.0 9.4

R-nw No 2 8.9, 8.9 8.9

C-F-nw

C

F

No

3 0, 0, 0† 0†

n.a.C-R-nwR

3 3.17, 2.91, 3.15 3.1

C-RW-nw* 2 3.26, 3.12 3.2

FS-lw

CFFT

lw

FYes 3

n.a.

9.6, 10.7, 10.6 10.3

F-lw No 2 10.2, 9.8 10.0

RS-lwR

Yes 3 9.6, 10.8, 10.1 10.2

R-lw No 2 11.1, 9.2 10.2

C-F-lw

C

F

No

3 0, 0, 0† 0†

n.a.C-R-lwR

3 5.8, 5.9, 5.9 5.9

C-RW-lw* 3 6.1, 5.9, 5.6 5.9*Cylinders were submerged in water at room temperature for a period equivalent to the total water-thawing time of the specimens under freezing and thawing cycles.†All cylinders completely disintegrated after about 100 cycles and could not be tested.Note: Avg. = average; C = plain concrete; CFFT = concrete-filled, fiber-reinforced-polymer tubes; F = freezing and thawing; lw = lightweight concrete; nw = normalweight concrete; n.a. = not applicable; R = room temperature; S = sustained loading; W = submerged in water. 1 ksi = 6.895 MPa.

January–February 2008 | PCI Journal4

Lafarge Canada Inc. designed and supplied the concrete. For the lightweight concrete, aggregates from industrial palletized slag of 3/8 in. (9 mm) maximum size were used. Because these aggregates have a high moisture absorption level (due to their high poros-ity), they were presoaked in water for 7 days before mixing. The mixture was proportioned to obtain a 3.15 in. (80 mm) slump. The average compressive strength at 28 days was 5.9 ksi (41 MPa).

For the normalweight concrete, commonly available normalweight aggregates of 3/4 in. (19 mm) maximum aggregate size were used, and the mixture had 3 in. (76 mm) slump. The 28-day compressive strength was 3.1 ksi (21 MPa). In both mixtures, no air entrainment

was used to simulate the worst possible situation for the concrete core under freezing and thawing conditions.

Fabrication process

Wooden bases were affixed to the bottoms of the twenty 6.57 in. × 13.15 in. (167 mm × 334 mm) GFRP tubes. The outer sides of the bases of the tubes were sealed using latex epoxy coring to prevent water drainage from the filled tubes. Plastic molds were used to cast the standard size 6 in. × 12 in. (150 mm × 300 mm) plain concrete cylinders. Both the GFRP tubes and the plastic molds were filled in three layers, and each layer was tapped 25 times, using the standard procedure, to ensure adequate consolidation. After 7 days, the wooden bases of CFFT specimens and plastic molds of plain concrete cylinders were removed and left to cure at room temperature (Fig. 4).

To obtain smooth concrete faces of the CFFT speci-mens for proper contact under sustained loading and also during testing at a later stage, a special grinding process was used. Each specimen was gripped by a chuck, leveled, and then ground with a diamond grinding disc applied to the face of the spinning specimen (Fig. 5). A 0.08 in. to 0.10 in. (2 mm to 3 mm) layer of concrete was removed from each end of each specimen as a result of this process.

Exposure and loading procedures

All specimens were first subjected to one load-ing cycle at room temperature up to a load level of approximately 30% of their respective ultimate confined strength

f '

cc at a rate of 0.012 in./min (0.3

mm/min) and then were unloaded immediately at the same rate. This was intended to simulate the effect of loading history, which may increase the moisture intake capacity of the concrete core if microcracks develop.

All specimens, including plain concrete cylinders, were then submerged in water for seven days, im-mediately before the application of the sustained loading and/or freezing and thawing cycles. This practice was adopted to increase the internal mois-ture content before freezing because the tight con-tact at the smooth interfaces of CFFT specimens (loaded in series) may hinder the internal moisture absorption of the concrete core during the thaw-ing phases of the cycles. Increasing the moisture content was intended to create conditions similar to those in field applications of CFFT piles in marine environments.

Figure 5. The specimen faces were ground to ensure a smooth face during loading.

Figure 4. After casting, test specimens were left to cure at room temperature.

5PCI Journal | January–February 2008

Four individual steel loading frames were built such that each accommodated the specimens of each group subjected to sustained loading in a concentric series. Each frame consisted of four 1.06-in.-diam-eter (27 mm) B16 threaded rods along with three 2-in.-thick (50 mm), 9.8 in. × 9.8 in. (250 mm × 250 mm), mild-steel square plates (Fig. 6). Two frames were placed in the cold room to accommodate speci-mens FS-nw and FS-lw, as shown in Fig. 7, and two frames were kept at room temperature to accommo-date specimens RS-nw and RS-lw.

The CFFT specimens were placed between two steel plates. A load cell and a hydraulic ram were placed concentrically behind one of the plates, and the third steel plate was then placed and anchored behind the ram to jack against. Once the desired load of 100 kip (445 kN), which produced 29% to 33% of ultimate strength, was reached, the nuts anchoring the middle steel plate were tightened, and the ram, the load cell, and the third plate were removed.

In two frames, the load cell was placed permanently

with the specimens between the two steel plates to monitor the load continuously. During the early days of the duration of exposure, the load tended to drop due to creep of concrete and relaxation of the threaded rods. The load was then increased back to the original loads as needed in all four frames to sustain the desired stress level in the specimens. The drops in load reduced tremendously after the

first two weeks (Fig. 8). For frames subjected to freezing and thaw-ing cycles, the temperature cycling also resulted in a 12% fluctua-tion of the sustained load level within each cycle (Fig. 9).

A special environmental chamber (cold room) was used to apply a freezing-and-thawing cycling regime in accordance with ASTM C666-97.13 This standard, which is for plain concrete, was used be-cause no standard exists for FRP-confined concrete. The objective

Figure 7. Pictured is a sustained loading setup of the frame in an environmental chamber.

Figure 6. This is a schematic view of the sustained loading setup. Note: CFFT = concrete-filled, fiber-reinforced-polymer tubes.

Hydraulic ram

Load cell CFFT specimens

Threaded rods

Steel plates

Increasing the moisture content was intended to create condi-tions similar to those in field applications of CFFT piles in

marine environments.

January–February 2008 | PCI Journal6

was to apply 300 cycles, where the concrete core temperature varied from +41 oF to -0.4 oF (+5 oC to -18 oC). In order to accomplish this pattern, the air temperature inside the chamber was varied from +50 oF (+10 oC) to -22 oF (-30 oC) within 5 hours and 25 minutes (Fig. 10).

Some specimens included thermocouples inside the concrete during casting to measure the core temperature, which helped determine the appropri-ate program to be used for controlling the environ-mental chamber. A tub was used to accommodate the sustained loading frames, which were frozen in air and thawed in water that was automatically pumped into the tub during the thawing period of each cycle and then drained. Specimens that were only subjected to freezing and thawing were also placed in the same tub.

Test setup and instrumentation

After the completion of the 300 freezing and thaw-ing cycles, all specimens were left to dry for at least 7 days until their core temperature reached that of the room. Each specimen was then instrumented to measure the axial strain using three displacement-type strain transducers with 4 in. (100 mm) gage length, spaced at 120 degrees around the perimeter at mid-height. Two 0.2 in. (5 mm) electric resistance strain gauges, spaced at 180 degrees, were used to measure the hoop strains at mid-height.

A 1124 kip (5000 kN) MTS testing machine test-ed the CFFT specimens under axial compression at a loading rate of 0.012 in./min (0.3 mm/min) (Fig. 11). In some of the CFFT specimens exposed to freezing and thawing only, it was noted that some localized concrete disintegration occurred at the ex-posed faces within a thin layer, which resulted in a rough concrete surface. For those specimens, a thin layer of cement paste was applied to ensure a flush surface with the GFRP tube edge during testing. No visual signs of distress were observed in the GFRP tubes.

Test results and discussion

Table 1 provides a summary of the ultimate strengths of all specimens. In the following sec-tions, the effects of various parameters and failure modes are discussed.

Figure 10. This plot shows how the temperature was varied with time through one freezing and thawing cycle.

-35

-30

-25-20

-15

-10

-5

05

10

15

0 1 2 3 4 5 6T i m e ( h r s )

Temp. ( ˚C) .Concrete core

Air

+59

+32

-31

Temp. (

oF)

Time, hour

Tem

pera

ture

, ℃

Tem

pera

ture

, ℉

Figure 8. This plot shows how the sustained load was varied with time at room temperature. Note: 1 kip = 4.45 kN.

0

1 0 0

2 0 0

3 0 0

4 0 0

5 0 0

0 5 1 0 1 5 2 0 2 5 3 0 3 5

Ti m e ( d a y s )

Force (kN)

lo a d in t e r v a ls

Load (kip)

Time, day

Load

, kip

Figure 9. This plot shows how the sustained load was varied with time during freezing and thawing cycles. Note: 1 kip = 4.45 kN.

0

100

200

300

400

500

5 10 15 20

Force (

kN)

Freeze

ThawOne cycle

Load (kip)

Time, hour

Load

, kip

7PCI Journal | January–February 2008

Plain concrete cylinders

The average compressive strength of C-R-nw specimens was 3.1 ksi (21.2 MPa), whereas that of the C-RW-nw was 3.2 ksi (22.0 MPa). Also, the average strength for both the C-R-lw and C-RW-lw specimens was 5.9 ksi (40.6 MPa). This suggests that the curing effect during the total water-thawing period was insignificant. Figure 12 shows the failed plain concrete cylinders.

More important, however, is the fact that all of the C-F-nw and C-F-lw plain concrete cylinders, which were exposed to freezing and thawing cycles, com-pletely crumbled after only 100 and 150 cycles, re-spectively, due to the lack of air entrainment (Fig. 13). As such, they could not be tested and it was clear that they had virtually zero residual strength.

CFFT specimens

Effects of freezing and thawing and sustained loading Figure 14 compares the axial strengths of all CFFT specimens, including normalweight and lightweight concrete specimens, under various conditions of exposure. Figures 15 and 16 give the typical stress-strain curves of various CFFT specimens. Figure 14 clearly shows that, relative to the control specimens at room temperature (R), no signs of any reduction in strength were evident. In fact, the normalweight specimens subjected to sustained loading at room temperature (R) demonstrated some increase in strength.

This strength increase was likely due to the creep effect of concrete, which resulted in additional axial compressive strains under the sustained load and

a tendency for additional radial tensile strains (dilation) that were restrained by the tube. As such, an active confinement pressure was likely generated prior to testing to failure.

Also, specimens subjected to the combined effects of freezing and thawing exposure and sustained loading (FS) experienced an

Figure 11. Shown is a test setup of a concrete-filled, fiber-reinforced-polymer tube specimen under axial compression.

Figure 12. Pictured are failure modes of plain concrete at room temperature. Note: C = plain concrete; lw = lightweight concrete; nw = normalweight concrete; R = room temperature.

January–February 2008 | PCI Journal8

were confined by identical GFRP tubes and exposed to identical conditions. Figure 17 shows a comparison between the average stress-strain responses of the normalweight and lightweight CFFT control specimens (R-nw and R-lw) relative to their respective unconfined strengths. The figure shows that both types reached a comparable ultimate confined strength

f '

cc , despite the large

difference in their f 'c values.

Figure 18 shows the normalized responses, in terms of the confinement ratio ( f '

cc / f '

c ), which

indicates that the lower-strength concrete (nor-malweight in this case) achieved a confinement ratio of about 3.0, whereas the higher-strength concrete (lightweight) achieved a ratio of only 1.7. This could be attributed to a lower dilation capac-ity, which agrees well with a previous study14 that suggested that medium- to high-strength concretes benefit less from FRP confinement than the low-strength concrete does due to their brittle nature and lower dilation capacity. This also explains the lower strength gain in lightweight concrete CFFT specimens subjected to sustained load and freezing and thawing exposure compared with their normal-weight counterparts, as mentioned previously.

Failure modes Figures 19 and 20 show sample failure modes of the normalweight and lightweight CFFT specimens after axial load testing, including both the control specimens at room temperature and the ones subjected to combined sustained loading and freezing and thawing cycles before testing. Regardless of the type of concrete and nature of exposure, failure consistently occurred due to fracture of the GFRP tube under a biaxial state of stress, including axial compressive and hoop tensile stresses. The most obvious form of failure was tensile fracture in the hoop direction; however, some evidence of crushing of the GFRP tube in the axial

even further increase in their axial strength, particularly for the normalweight specimens. This effect was likely attributed to the fact that the volume increase of the concrete core due to expansion of the frozen moisture was restrained in three directions, in addition to the creep effect indicated earlier. The two combined effects resulted in a triaxial state of stress, essentially activating the GFRP tube and providing a state of active confinement, even before testing the specimens, which was analogous to a prestressing effect.

In specimens subjected to freezing and thawing (F) only, this effect did not occur because the concrete was relatively free to expand longitudinally, allowing most of the internal pressure to be relieved. Also, in lightweight concrete, no remarkable increases in strength were noticed, as will be discussed in the following section.

Combined effects of concrete strength and density In this study, the unconfined compressive strengths

f '

c of the normalweight

and lightweight concrete were different, 3.2 ksi and 5.9 ksi (22 MPa and 41 MPa), respectively, yet both types of concrete

Figure 13. Pictured are failure modes of plain concrete subjected to freezing and thawing. Note: C = plain concrete; F = freezing and thawing; lw = lightweight concrete; nw = normalweight concrete.

Figure 14. The graph shows a strength comparison of concrete-filled, fiber-reinforced-polymer tube specimens under various conditions. Note: F = freezing and thawing; lw = light-weight concrete; nw = normalweight concrete; R = room temperature; S = sustained load; f'c = unconfined compressive strength. 1 ksi = 6.895 MPa.

'cf Lightweight ( = 5.9 ksi)

'cf11

9

10

8

7

6

5

4

3

2

1

0

FS

-nw

F-n

w

RS

-nw

R-n

w

FS

-lw

F-l

w

RS

-lw

R-l

w

Con

fined

str

engt

h

,

ksi

'ccf

Normalweight ( = 3.2 ksi)

9PCI Journal | January–February 2008

direction was also noticed in several specimens. Failure was generally accompanied by a complete loss of load capacity. In a few cases, however, a small drop in load was observed first due to gradual fracture of the tube (Fig. 16). The load then increased again until the tube completely failed.

Summary and conclusions

CFFTs were subjected to various loading and/or exposure conditions, including sustained load-ing, freezing and thawing cycles, or a combination of the two. After exposure to various conditions the specimens were tested to failure under axial compression. Based on the results of these tests, the following conclusions were drawn.

CFFT specimens survived all types of exposure without any noticeable reduction in their ultimate compressive strengths. Plain concrete cylinders,

Figure 15. These graphs show stress-strain curves of normalweight specimens. Note: F = freezing and thawing; nw = normalweight concrete; R = room temperature; S = sustained load. 1 ksi = 6.895 MPa.

-0 .03 -0.02 -0.01 0 0.01 0.02 -0 .03 -0 .02 -0 .01 0 .01 0 .02

11109876543210

Strain

Str

engt

h, k

si

11109876543210

Strain

Str

engt

h, k

si

FS-nw

F-nw

RS-nw

R-nw

Axial Circumferential Axial Circumferential

Figure 16. These graphs show stress-strain curves of lightweight specimens. Note: F = freezing and thawing; lw = lightweight concrete; R = room temperature; S = sustained load. 1 ksi = 6.895 MPa.

-0 .018 -0 .012 -0 .006 0 0 .006 0 .012 -0 .018 -0 .012 -0 .006 0 0 .006 0 .012

11109876543210

Strain

Str

engt

h, k

si11109876543210

Strain

Str

engt

h, k

si

FS-lw

F-lwRS-lw

R-lw

Axial Circumferential Axial Circumferential

Figure 17. This graph shows the influence of concrete strength and density on confine-ment ratio on stress-strain response. Note: lw = lightweight concrete; nw = normalweight concrete; R = room temperature; f'c = unconfined compressive strength. 1 ksi = 6.895 MPa.

-0 .0 1 5 -0 .0 1 -0 .0 0 5 0 0 .0 0 5 0 .0 1

' 3.2ksicf =

Normalweight

(R-nw)

' 5.8ksicf =

Lightweight

(R-lw)

Strain

Str

ess

, ksi

Axial Circumferential

1 1

1 0

9

8

7

6

5

4

3

2

1

0

1 1

1 0

9

8

7

6

5

4

3

2

1

0

' ccf

January–February 2008 | PCI Journal10

however, completely disintegrated, with virtually no residual strength, after 100 to 150 freezing and thawing cycles due to the lack of air entrainment.

In CFFT specimens containing normalweight, low-strength concrete, the sustained loading resulted in a noticeable increase in the axial strength. This be-havior is likely due to creep effects, which tended to increase the axial and radial concrete strains and therefore induced active confinement by means of the restraint imposed by the tube. Also, freez-ing and thawing exposure under sustained loading resulted in even further increases in axial strength. The expansion of the frozen concrete core, which was restrained in all directions, likely generated internal pressure that added to the active confine-ment pressure induced by creep. This behavior was not noticed in CFFT specimens with lightweight,

Figure 19. Pictured are the failure modes of normalweight concrete-filled, fiber-reinforced-polymer tubes. Note: F = freezing and thawing; nw = normalweight concrete; R = room temperature; S = sustained load.

Figure 20. Pictured are the failure modes of lightweight concrete-filled, fiber-reinforced-polymer tubes. Note: F = freezing and thawing; lw = lightweight concrete; R = room temperature; S = sustained load.

Figure 18. Shown is the influence of concrete strength and density on confinement ratio on normalized responses. Note: f'c = unconfined compressive strength; f'cc = confined compres-sive strength; lw = lightweight concrete; nw = normalweight concrete; R = room tempera-ture. 1 ksi = 6.895 MPa.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

-8 -6 -4 -2 0

Normalweight(R-nw)

Lightweight(R-lw)

Normalized axial strain ( )'' / cccε ε

' ccf' cf

)(

11PCI Journal | January–February 2008

References

1. Mirmiran, A., and M. Shahawy. 1997. Behavior of Concrete Columns Confined by Fiber Composites. Journal of Struc-tural Engineering (May): pp. 583–590.

2. Fam, A., R. Greene, and S. Rizkalla. 2003. Field Applica-tions of Concrete-Filled FRP Tubes for Marine Piles. In SP-215: Field Application of FRP Reinforcement: Case Stud-ies, pp. 161–180. Farmington Hills, MI: American Concrete Institute (ACI).

3. Fam, A., M. Pando, G. Filz, and S. Rizkalla. 2003. Precast Composite Piles for the Route 40 Bridge in Virginia Us-ing Concrete-Filled FRP Tubes. PCI Journal, V. 48, No. 3 (May–June): pp. 32–45.

4. Fam, A., and S. Rizkalla. 2003. Large Scale Testing and Analysis of Hybrid Concrete/Composite Tubes for Circular Beam-Column Applications. Construction and Building Materials, V. 17, No. 6–7: pp. 507–516

5. Fam, A., B. Flisak, and S. Rizkalla. 2003. Experimental and Analytical Investigations of Concrete-Filled Fiber-Reinforced Polymer Tubes Subjected to Combined Bending and Axial Loads. ACI Structural Journal, V. 100, No. 4: pp. 499–509.

6. Naguib, W., and A. Mirmiran. 2002. Time-Dependent Behavior of Fiber-Reinforced Polymer Confined Concrete Column under Axial Loads. ACI Structural Journal, V. 99, No. 2: pp. 142–148.

7. Soudki, K. A., and M. F. Green. 1997. Freeze Thaw Durabil-ity of Compression Members Strengthened by Carbon Fiber Wrapping. Concrete International, V. 19, No. 8: pp. 64–67.

8. Toutanji, H., and F. Rey. 1998. Performance of Concrete Columns Strengthened with Advanced Composites Subjected to Freeze-Thaw Conditions. In Proceedings from the 1998 CDCC Conference, pp. 351–360.

9. Karbhari, V. M. 2002. Response of Fiber Reinforced Con-fined Concrete Exposed to Freeze-Thaw Regimes. Journal of Composites for Construction, V. 6, No. 1: pp. 35–40.

10. Kong, A., A. Fam, and M. F. Green. 2005. Freeze-Thaw Behavior of FRP-Confined Concrete under Sustained Loads. In SP-230: 7th International Symposium on Fiber-Reinforced (FRP) Polymer Reinforcement for Concrete Structures, pp. 705–722. Farmington Hills, MI: ACI.

11. Fam, A. Z., D. Schnerch, and S. H. Rizkalla. 2002. Moisture Effect on Durability of Concrete-Filled GFRP Tubular Piles. RD-02-05. North Carolina State University Constructed Facilities Laboratories.

medium-strength concrete.

The GFRP tube confinement effectiveness was substantially higher for normalweight, low-strength concrete compared with the lightweight, medium-strength concrete. The dilation capacity of medium-strength concrete, and therefore the effectiveness for engaging the GFRP tube, is likely lower than that of low-strength concrete.

In previous studies, precast CFFT piles (without any internal reinforcement) have been shown to be quite reliable under a variety of structural loading condi-tions,1,4,5,6,14 including several field applications,2,3 and also as prestressed piles.15 This paper has dem-onstrated that commercially available GFRP tubes (used in pipeline industries) with fibers oriented in the longitudinal and circumferential directions are quite reliable, not only as a structural form for concrete, but also as protective jackets, under severe freezing and thawing combined with sustained load-ing conditions. The study also confirms an earlier finding14 that the tubes could be filled with low-strength concrete, taking advantage of their superior confinement effectiveness in this case.

Additional research is needed to include the effects of flexure or eccentric loading, which induce substantial cracking on freezing and thawing behavior. Also, the effect of concrete density on the behavior could not be distinctly verified in this study because of the difference in concrete strength of the lightweight and normalweight concretes. It is suggested that CFFT specimens with lightweight and normalweight concretes of the same aggregate size and compressive strength should be tested and compared.

Acknowledgments

The authors acknowledge the financial support pro-vided by the Network of Centres of Excellence on Intelligent Sensing for Innovative Structures (ISIS Canada), the Natural Sciences and Engineering Research Council of Canada, Lancaster Composite, and the University of Sherbrooke. The authors are also grateful to Dave Tryon, Paul Thrasher, Neil Porter, and Jamie Escobar of Queen’s University. The authors thank the PCI Journal reviewers and editors for their comments and suggestions.

January–February 2008 | PCI Journal12

15. Fam, A., and S. Mandal. 2006. Prestressed Concrete-Filled Fiber-Reinforced Polymer Circular Tubes Tested in Flexure. PCI Journal, V. 51, No. 4 (July–August): pp. 42–54.

Notation

f 'c

= unconfined compressive strength of concrete

f 'cc

= ultimate confined strength of concrete

12. Fiberglass Composite Pipe Group. Series 2000 Fiberglass Pipe and Fittings data sheet. Ameron International: Houston, TX.

13. American Society of Testing and Material (ASTM) Subcom-mittee C09.67. 1997. Standard Test Method for Resistance of Concrete to Rapid Freezing and Thawing. ASTM C666-97, pp. 314–319. West Conshohocken, PA: ASTM.

14. Mandal, S., A. Hoskin, and A. Fam. 2005. Influence of Con-crete Strength on Confinement Effectiveness of Fiber-Rein-forced Polymer Circular Jackets. ACI Structural Journal, V. 102, No. 3: pp. 383–392.

About the authors

Amir Fam, Ph.D., P.Eng., is an associate professor and Canada Research Chair in Innovative and Retrofitted Structures for the Department of Civil Engineering at Queen’s University in Kingston, ON, Canada.

Andrew Kong, M.Sc., is Project Engineer and Restoration Specialist in Carl Walker Inc., Cherry Hill, New Jersey, and a former graduate student at the Department of Civil Engineering at Queen’s University in Kingston.

Mark F. Green, Ph.D., P.Eng., is a professor for the Department of Civil Engineering at Queen’s University in Kingston.

Synopsis

The concept of precast concrete-filled fiber-reinforced polymer tubes (CFFTs) has been introduced as a viable alternative to conventional reinforced concrete members, particularly in structural applications in harsh marine environments. The prefabricated composite tube in the CFFT system resists corrosion and protects the concrete core from various deterioration and damage mecha-nisms, while eliminating the need for steel reinforcement and providing excellent confinement to the concrete core.

This paper reports the findings of an experimental investigation into the structural response of CFFTs subjected to combined sustained axial loads and freezing and thawing cycles, before being tested to failure under axial compression. CFFT control specimens subjected to either sustained load only or freezing and thawing cycles only were also tested.

The study found that the CFFTs can survive all types of exposure without any noticeable reduction in strength, whereas concrete without the GFRP tubes exposed to similar conditions will completely disintegrate in a short time. Sustained loading could increase the strength of the concrete due to the restrained expansion of the concrete by the tubes, which induces active confinement before loading. It was also found that confinement was more effective for the lower-strength concrete than it was for the higher-strength concrete due to the fact that it has more to gain from confinement than a concrete that is already relatively strong.

Keywords

Composites, confinement, durability, fiber-reinforced polymers, freezing, glass fiber-reinforced polymers, jackets, piles, sustained loads, thawing.

Review policy

This paper was reviewed in accordance with the Precast/Prestressed Concrete Institute’s peer-review process.

Reader comments

Please address any reader comments to PCI Journal editor-in-chief Emily Lorenz at elorenz@pci.org or Pre-cast/Prestressed Concrete Institute, c/o PCI Journal, 209 W. Jackson Blvd., Suite 500, Chicago, IL 60606. J