Construction and Building

Construction and Building Materials 21 (2007) 118125

MATERIALS
www.elsevier.com/locate/conbuildmat

Using bres to enhance the properties of concrete columns

M.N.S. Hadi
*
School of Civil, Mining and Environmental Engineering, University of Wollongong, NSW 2522, Australia Received 20 January 2005; received in revised form 16 June 2005; accepted 30 June 2005 Available online 19 August 2005

Abstract This paper explores the eects of adding synthetic reinforcing bres to high-strength reinforced concrete columns and in particular only to the cover of the columns. An experimental program was conducted where seven circular reinforced concrete columns were tested with varying bre content one contained no bres, two contained bres throughout the cross-section and four contained bres only in the outer concrete. The other column properties were kept the same for all the seven columns. All seven columns were tested by the application of a concentric, axial compression force. It was found that although only minor improvements were noticeable for a bre content of 0.1%, the addition of 0.3% polypropylene bres increased the load at which cover spalling took place. It was also found that the columns containing both brous high strength concrete (FHSC) in the outer concrete and HSC in the core exhibited higher levels of ductility than the columns containing FHSC throughout the entire cross-section. 2005 Elsevier Ltd. All rights reserved.
Keywords: Fibre; RC columns; Ductility

1. Introduction Over the past 20 years, research of and improvements to concrete mix design have resulted in an increase of three to fourfold in available concrete strengths. With respect to columns, a larger compressive strength means a smaller cross-sectional area required, resulting in better utilisation of available space and materials. These higherstrength columns, however, have been shown to contain weaknesses. As the compressive strength increases so too does the brittleness and while increasing the amount of lateral reinforcement reduces this brittleness, it also increases the columns susceptibility to early cover spalling [1]. In order for high-strength concrete (HSC) columns to be eective and superior to normal strength concrete columns, these weaknesses need to be overcome. Early spalling of the cover concrete arises due to the connement eect provided by the reinforcement. When
*

Tel.: +61 2 4221 4762; fax: +61 2 4221 3238. E-mail address: mhadi@uow.edu.au.

a column is axially compressed, material is pushed outwards resulting in increased cross-section dimensions. The core concrete is conned by reinforcement, but the unconned cover concrete outside the reinforcement continues to be pushed outwards, placing it in tension, and forcing it to separate from the core concrete. Once this cover has spalled the column has less cross-sectional area and hence has a reduced load-carrying capacity. Research has shown that the addition of bres to the concrete helps to arrest the onset of early cover spalling [13]. This research has also shown that ductility of the HSC columns can be improved through the addition of bres. One of the problems associated with using bres is that the bres may not be uniformly distributed throughout the member. In particular, bres may clump together in the core of the column, thus the density and compressive strength at that section may be reduced. These problems may be overcome by locating bres only in the cover concrete, to prevent early cover spalling whilst having the core of the column containing only

0950-0618/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2005.06.028

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125

119

non-brous concrete to maintain sucient compressive strength. This process, although laborious, would result in more ecient use of materials. This study of bre reinforced concrete columns investigates a new method of column construction which results in brous high strength concrete (FHSC) being located only in the cover concrete while plain HSC is located in the remaining core concrete. It is proposed that this new type of column will perform in a superior manner to columns which contain FHSC throughout the entire cross-section as problems such as the movement of bres towards the centre of the column, away from the cover, during vibration will be overcome. This new type of column also provides a more ecient use of materials as bres are located in the cover to prevent early cover spalling, while the remainder of the column contains only plain HSC so the integrity and density of the core remain unaected. The aim of this study is to nd a more ecient design for bre reinforced high strength concrete columns.

testing, that as the amount of tie reinforcement increased, the ultimate column strength (or the load at which early cover spalling took place) decreased while the ductility of the column increased. Liu et al. [8] conducted tests on 12 high strength concrete columns to investigate the issue of early cover spalling. It was found that the column strength was the greater of the column spalling load and the capacity of the conned core and that the study provided further evidence that the design load can be taken as 0.85 times the capacity of the entire concrete section. 2.3. Fibres within concrete columns Adepegba and Regan [9] carried out tests on steel bre reinforced concrete columns. It was found that the addition of steel bres at any of the tested bre contents did not increase the ultimate load of the column. It was noted, however, that the experiments did not investigate the post-failure behaviour improvements gained by the addition of steel bres. Ganesan and Murthy [10] studied the eect of varying the amount of lateral reinforcement on steel bre-reinforced and non-bre-reinforced concrete columns. It was found that as the amount of lateral reinforcement increased, larger strength increases were obtained from the brous columns compared with the non-brous columns and the addition of steel bres resulted in better strength and ductility. Campione [11] presented a mathematical model which determined the stressstrain relationship for bre reinforced concrete columns. The analytical expressions allowed the determination of the maximum strength and strain capacity of circular or square, high-strength or normal-strength, brous or non-brous concrete columns. Explanations were also given regarding the region of the column cross-section which could be considered eectively conned by the reinforcement. It was also noted that the model was veried through experimental testing. Foster [2] presented a design model for calculating the ductility of bre reinforced high strength concrete columns. It was demonstrated through design examples that the spacing of reinforcement ties could be increased with the inclusion of steel bres in the concrete for a given level of ductility. Sarker [3] studied the eect of adding synthetic bres to high strength concrete columns. It was also proposed that longer bres at a higher percentage content would produce better column performance.

2. Literature review 2.1. High strength concrete columns Foster and Attard [4] tested 68 concrete columns by varying load eccentricities, concrete strengths and the amount of reinforcement. It was found that increasing the amount of longitudinal reinforcement and/or decreasing tie spacings will reduce the ultimate capacity for combined bending and compression for all concrete strengths, as the reinforcement formed a natural plane of separation between the core and cover concrete. Pessiki and Pieroni [5] conducted studies on eight spirally reinforced concrete columns. The study found that ductility decreased with increasing concrete strength and the columns with higher amounts of longitudinal reinforcement maintained peak load for longer but displayed less ductility than the columns with less longitudinal reinforcement. Razvi and Saatcioglu [6] studied the eect of varying reinforcement on 20 high-strength concrete columns. It was found that high-strength concrete columns displayed extremely brittle behaviour unless conned by suitable transverse reinforcement. Also, the presence of longitudinal reinforcement improved the performance of the columns due to a better conning eect. 2.2. Early cover spalling in high strength concrete columns Foster et al. [7] used nite element modelling to investigate the eect of cover spalling for both high-strength (>60 MPa) and normal-strength concrete columns subject to concentric compression loads. It was found through the model, and veried through experimental

3. A new casting technique for concrete columns As discussed above, bres increase ductility and also help to arrest the onset of early cover spalling (ECS). As

120

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125 Table 1 Column number by bre content and location Col C1 C2 C3 C4 C5 C6 C7 Fibre content (%) 0 0.1 0.1 0.1 0.3 0.3 0.3 Fibre location N/A Entire cross-section Cover + extending into core (outer 45 mm) Cover only (outer 22.5 mm) Entire cross-section Cover + extending into core (outer 45 mm) Cover only (outer 22.5 mm)

the column cross-section increases in size due to Poisson expansion, inward forces are provided to the core by the lateral reinforcement and the core concrete is eectively conned. The cover concrete on the other hand has no connement and hence is forced outwards. This eect greatly contributes to ECS as the tensile bond between the core and the cover concrete is relatively weak and separation is inevitable. The addition of either steel or polypropylene bres has shown to increase the tensile strength of the concrete and assist with the prevention of ECS. 3.1. Uncertainties using bres in HSC One of the disadvantages of using bres in concrete columns is that the bres are not always distributed evenly throughout the column. As internal vibration takes place, the bres tend to be drawn towards the centre of the column. This means that the core concrete contains a larger percentage of bres than was initially added to the concrete mix whilst the outside layer of concrete contains a lower percentage of bres. The excess bres in the core concrete would lower the density, reducing the compressive strength of the column and the lack of bres in the cover concrete would mean that ECS would not be prevented. In order to overcome the uncertainties outlined above, it was necessary to trial a new column crosssection. This cross-section involved the use of brous concrete only where it was required and ensured that bres were not vibrated away from the regions where they were needed. The proposed column cross-section, developed during the course of this study, was constructed using plain HSC in the core concrete and FHSC in the cover concrete only. A secondary trial cross-section was also proposed which involved the cover concrete containing FHSC which extended through the conventional steel reinforcement into the core. The reason for the secondary cross-section was to ensure the cover concrete was tied into the core concrete as it was thought the cross-section which was proposed initially may promote ECS due to the method of construction. These two cross-sections eectively reduced uncertainty about the location of the bres and increased condence that bres were present within the cover and were not present in the core.

6N12 925 mm 205 mm

Fig. 1. Columns reinforcement details.

R10@50 mm

4. Experimental program Selection of the optimum bre quantity was critical to this study. Firstly, it was important to construct a column with 0% bres (Column C1) to provide a reference point to which results could be compared. Secondly, it was decided that the recommended manufacturers dosage of 0.9 kg/m3 (0.1% bres) should be used to provide

information as to whether the minimum recommended dosage for shrinkage was sucient for strength and ductility (Columns C2, C3 and C4). Finally, as for the amount of bres which would be sucient for this study, 0.1% was considered the minimum for strength and ductility purposes and 0.5% was considered the maximum for placement purposes. Hence, a value between these two of 0.3% bres (2.7 kg/m3) was chosen (C5, C6 and C7). In total seven columns were cast and tested. From each batch of bres three columns were proposed. One containing a uniform brous cross-section (C2 and C5), one containing bres in the cover extending into the core (C3 and C6) and one containing bres located only in the cover (C4 and C7). The bre content and location for each column are shown in Table 1 and details and diagrams of the columns are presented in the following sections. The reinforcement of all columns consisted of 6N12 bars (12 mm diameter deformed bars with 500 MPa tensile strength) and R10 (10 mm plain bars with 250 MPa tensile strength) at 50 mm helices. Details of the column specimens are shown in Fig. 1. 4.1. Casting procedure The high strength concrete (HSC) was placed into two wheelbarrows so casting of Column C1 could begin. In order to create the brous high strength concrete

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125

121

(FHSC) the HSC was placed directly from the concrete chute into the concrete mixer. The bres were added evenly to the concrete by hand as it descended the chute. The mixing drum was lled to the correct predetermined height so that a total volume of 0.125 m3 was present. The mixer was switched on for 1.52 min until the bres were dispersed evenly throughout the mix. The quantities of bres added to the concrete were 0.11 kg for the 0.1% mix and 0.34 kg for the 0.3% mix. In order to determine the properties of both the HSC and FHSC, sample cylinders and beams were cast. For each of the three batches (0%, 0.1% and 0.3%) six small cylinders were made for compressive tests (three for 7-day strength and three for 28-day strength), three large cylinders were made for indirect tensile tests at 28 days, and three beams were made for exural tests at 28 days. Each of the samples was cleaned and lubricated prior to casting. For the columns which had uniform cross-sections, pouring and vibrating were conducted in three stages. The concrete was scooped into the columns for 1/3 of the height before being vibrated with an electric vibrator. The middle and top 1/3 were poured in the same manner and nally the surface was nished with a wet trowel. No problems were encountered in pouring Column 1 (0%), Column 2 (0.1%) or Column 5 (0.3%). The columns which were to have a smaller HSC core (Column C3 and Column C6) were constructed with a PVC pipe located inside the steel reinforcement (see Fig. 2). The HSC was scooped into the PVC pipe whilst the FHSC was placed around the outside of the PVC pipe. The FHSC was pushed with 6 mm bars and the sides of the formwork were tapped with a mallet to ensure the FHSC dropped to the bottom of the column. The column was lled in this way until the HSC inside the PVC pipe and the FHSC outside the PVC pipe were equal and at roughly 1/3 the column height (see Fig. 3(a)). The PVC pipe was then lifted until it was

roughly 50 mm inside the concrete (see Fig. 3(b)). The vibrator was then placed down the centre of the PVC pipe to vibrate the core whilst the cover was rodded with 6 mm bars and tapped with a mallet. This process was then repeated for the middle and upper 1/3 of the column (see Fig. 3(c)(e)) and the top surface was nished with a wet trowel. The columns which were to have a larger HSC core, or bres only located in the cover (C4 and C7) were constructed with a Perspex sheet fabricated into a tube with diameter 163 mm located outside, but in contact with, the steel reinforcement (see Fig. 4). The HSC was placed in the centre of the Perspex tube whilst the FHSC was placed around the outside of the tube in the gap between the Perspex and the formwork. Although the gap in which to place the FHSC was smaller than the gap for the PVC pipe, the concrete dropped to the bottom of the column with ease as there were no obstructions and pushing with 6 mm bars was not required. The column was lled in this way until the HSC inside the Perspex tube and the FHSC outside the Perspex tube

Fig. 3. Column casting procedure.

Fig. 2. Columns 3 and 6 small HSC core.

122

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125

Fig. 4. Columns 4 and 7 large HSC core.

were equal and at roughly 1/3 the column height (see Fig. 3(a)). The Perspex tube was then lifted until it was roughly 50 mm inside the concrete (see Fig. 3(b)). The vibrator was then placed down the centre of the Perspex tube to vibrate the core whilst the cover was rodded with 6 mm bars and tapped with a mallet. This process was then repeated for the middle and upper 1/3 of the column (see Fig. 3(c) and (d)), although the upper 1/3 was not vibrated immediately. The nal 40 mm was not placed straight away (see Fig. 3(e)) in order to allow for the strain gauge wires to be threaded through the hole in the side of the formwork. Once this was carried out, the Perspex tube was replaced in the column, the nal 40 mm was cast and vibration of the top 1/3 took place. Finally, the top surface was nished with a wet trowel. At the conclusion of casting, all specimens were covered with wet Hessian and plastic sheets to prevent moisture loss. The cylinder and beam samples were stripped and placed in a curing tank whilst the columns were stripped after 7 days and placed under wet Hessian and covered with plastic sheets. 4.2. Column testing In order to test the columns the following procedure was employed. The columns were removed from the curing environment and each was weighed and measured (the recorded values were an average of two height measurements and an average of two diameter measurements at mid height). Each column was capped at the base with high-strength plaster and once the plaster had set, the column was inverted, lifted into the testing machine and capped at the opposite end, under a preload of around 14 kN provided by the testing machine. While the high-strength plaster was allowed to set a galvanised safety cage was placed around the column to minimise any potential damage caused by ying

debris. Once sucient time was allowed for the plaster to set, testing of the column took place. Loading of the column was controlled by displacement, rather than by load, so that post-peak performance could be monitored. The rate of loading was varied depending on the response of the column but a value of 0.3 mm/min was found to be ideal for speed and accuracy. For the testing of C5, C6 and C7, where testing was required to be repeated, the rate was increased to between 1.0 and 3.0 mm/min. During testing, deection readings were taken by hand every 20 kN so that loaddeection curves could be created.

5. Results 5.1. Compressive strength of concrete Compressive strength testing was carried out after 7 days and after 28 days. For each of the three batches (0%, 0.1% and 0.3%), three 100 mm 200 mm cylinders were tested. A rubber cap was used for the nine cylinders tested after 7 days while a high-strength plaster cap was used for the nine cylinders tested after 28 days. The results for the 7-day tests are shown in Table 2 and the 28-day test results can be seen in Table 3. 5.2. Indirect tensile strength of concrete Indirect tensile testing was used to determine the tensile strength increase due to the presence of bres in the high strength concrete. The 150 mm 300 mm cylinders were tested after 28 days, and results can be seen in Table 4. 5.3. Flexural strength of concrete Modulus of rupture tests were used to determine the exural strength increase due to the presence of bres in

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125 Table 2 Compressive strength at 7 days Batch M0 0.0% Sample mark M0-D M0-E M0-F M1-D M1-E M1-A M3-D M3-E M3-F Compressive strength (MPa) 36.2 38.9 38.5 43.8 42.2 46.9 40.0 40.8 43.7 Compressive strength avg. (MPa) 37.9 M0 0.0% 44.3 M1 0.1% 41.5 M3 0.3% M0-A M0-B M0-C M1-A M1-B M1-C M3-A M3-B M3-C Table 5 Flexural strength Batch Sample mark Tensile strength (MPa) 4.8 6.1 5.0 5.3 4.2 4.4 5.0 5.0 5.0 Tensile strength avg. (MPa) 5.3

123

Relative strength to batch M0 1.00

M1 0.1%

4.6

0.88

M3 0.3%

5.0

0.95

Table 3 Compressive strength at 28 days Batch bre content M0 0.0% Sample mark M0-A M0-B M0-C M1-B M1-C M1-F M3-A M3-B M3-C Ult. comp. strength (MPa) 54.3 64.9 67.4 67.4 69.0 54.2 68.3 60.3 66.7 Comp. strength avg. (MPa) 62.2 Relative strength to batch M0 1.00 C1 C2 C3 C4 C5 C6 C7 Table 6 Results of testing the columns Col Yield load (kN) 1580 1780 1660 1520 1725 1937 1774 Displ. at yield load (mm) 3.6 2.9 3.3 4.2 3.7 4.1 4.0 Ult. load (kN) 1780 1849 1830 1669 1935 2009 1885 Displ. at ult. load (mm) 9.5 3.0 9.2 9.0 5.1 4.8 8.9 Max displ. (mm) 14.8 11.2 13.6 >19.5 >78.0 >39.0 28.5

M1 0.1%

63.5

1.02

M3 0.3%

65.1

1.05

Table 4 Indirect tensile strength Batch bre content M0 0.0% Sample mark M0-A M0-B M0-C M1-A M1-B M1-C M3-A M3-B M3-C Tensile strength (MPa) 6.2 7.3 7.4 7.6 7.8 6.0 7.4 6.9 7.6 Tensile strength avg. (MPa) 7.0 Relative strength to batch M0 1.00

M1 0.1%

7.1

1.02

M3 0.3%

7.3

1.04

the seven tested columns. Ductility of each of the tested columns was calculated and are shown in Table 7. These ductilities were calculated as a ratio of the ultimate displacement to the yield displacement as one measure and as the ratio of the maximum displacement to the yield displacement as a second measure. The ductility of Column C1 was used as a reference value for the remaining columns. In order to assess the benets of varying the bre content, C1, C2 and C5 were compared. All three columns had uniform cross-sections with bre contents of 0.0%, 0.1% and 0.3%, respectively. Fig. 5 shows that cover spalling in both C2 and C5 occurred at a higher load than C1 indicating better performance due to the

the high strength concrete. The 100 mm 100 mm 500 mm beams were tested after 35 days and results can be seen in Table 5. While all specimens failed within the middle third of the span (indicating a exural failure), the specimens from Batch M0 (0.0%) separated whilst the specimens from Batches M1 (0.1%) and M3 (0.3%) remained intact. 5.4. Results of testing the columns Table 6 shows yield load, corresponding displacement, ultimate load and the maximum displacement of

Table 7 Calculated ductility of the tested columns Col Ductility: ult. disp./ yield disp. 2.7 1.1 2.8 2.2 1.4 1.2 2.2 Ductility relative to C1 1.0 0.4 1.0 0.8 0.5 0.4 0.8 Ductility: max. disp./ yield disp. 4.2 3.9 4.1 >4.6 >21.1 >9.5 7.1 Ductility relative to C1 1.0 0.9 1.0 >1.1 >5.1 >2.3 1.7

C1 C2 C3 C4 C5 C6 C7

124
2500 2000

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125

2500 2000

Load (kN)

Load (kN)

1500 1000 500 0 0 10 20 30 40

1500 1000 500 0 0 10 20 30 40

Deflection (mm)
C1 C2 C5

Deflection (mm)
C4 C7

Fig. 5. Columns C1, C2 and C5 varying bre content (bres located throughout entire cross-section).

Fig. 7. Columns C4 and C7 varying bre content (bres located in cover only).

presence of bres. The percentage of bres present also had an eect as cover spalling for C5 occurred at a higher load than for C2. The eect of varying the percentage of bres on cover spalling can also be shown through the comparison of C3 and C6. The columns contained both high-strength concrete (HSC) and brous, high-strength concrete (FHSC) in the same locations but C3 contained 0.1% FHSC while C6 contained 0.3% FHSC. Fig. 6 shows C6 reached a load of 2000 kN before cover spalling took place while C3 only reached a load of 1660 kN again indicating that increased bre content yields better results. Column C4 and Column C7 may also be compared in a similar manner as both contained bres in identical locations but Column C4 contained 0.1% FHSC while Column C7 contained 0.3% FHSC. Fig. 7 show the cover spalling load of 1840 kN for Column C7 was higher than the cover spalling load of 1570 kN for Column C4 again indicating that high strength concrete containing 0.3% bres performs much better than high strength concrete containing only 0.1% bres. In order to assess the benets of varying the location of the bres, Columns C2, C3 and C4 were compared. All three columns contained FHSC from

the same batch but C2 consisted of a uniform crosssection with FHSC throughout, C3 contained FHSC in the cover concrete extending through the region of the steel reinforcement into the core which contained plain HSC and C4 contained FHSC only in the cover concrete, i.e., only the outer 22.5 mm, and HSC throughout the remainder of the cross-section. As can be seen in Fig. 8, cover spalling took place at a higher load for C2 than for C3 and C4 indicating that reducing the amount of bre present in the crosssection reduces the load at which cover spalling takes place for a low bre content of 0.1%. Columns C5, C6 and C7 contained bres located in identical regions to C2, C3 and C4, respectively, but at three times the bre dosage, i.e., 0.3%. However, as shown by Fig. 9, varying the location of the bres did not have a signicant eect on the load at which cover spalling took place. Cover spalling of C7 occurred at a slightly lower load than that for C5 but cover spalling of C6 occurred at a slightly higher load than that for C5. It was concluded that at the 0.3% bre content, all of the three dierent cross-sections performed in a similar manner, regarding cover spalling, indicating that selection of a cross-section containing FHSC in the

2500

2500 2000

Load (kN)

Load (kN)

2000 1500 1000 500 0 0 10 20 30 40

1500 1000 500 0 0 10 20 30 40

Deflection (mm)
C3 C6

Deflection (mm)
C2 C3 C4

Fig. 6. Columns C3 and C6 varying bre content (bres located in cover plus extending into the core).

Fig. 8. Columns C2, C3 and C4 varying bre location (constant bre content of 0.1%).

M.N.S. Hadi / Construction and Building Materials 21 (2007) 118125

2500 2000

125

7. Conclusions This study of bre reinforced concrete columns involved the testing of seven high strength concrete columns with varying bre contents and bre locations. The main aim of this study was to determine whether placing brous concrete only in the outer cover, would be sucient to maintain, or enhance, the desirable properties of high strength concrete columns. The following is a list of the conclusions drawn from this study  Introducing 19 mm polypropylene bres to the concrete increases the load at which cover spalling takes place.  Increasing the bre content from 0.1% to 0.3% results in higher strengths being reached before cover spalling takes place.  The method used to place the bres in the cover is new and was found to be very eective for casting the columns but laborious.  Placing the bres in the cover only ensured that the bres were evenly distributed, the integrity of HSC core was maintained and the tendency of the bres to vibrate away from cover into the core was reduced.  Placing 0.3% bres only in the cover concrete did not reduce the load at which cover spalling took place.  Placing 0.3% bres in both the cover concrete and through the steel reinforcement into the core increases the load at which cover spalling takes place.

Load (kN)

1500 1000 500 0 0 10 20 30 40

Deflection (mm)
C5 C6 C7

Fig. 9. Columns C5, C6 and C7 varying bre location (constant bre content of 0.3%).

cover and plain HSC in the core, would result in a more ecient use of materials.

6. Discussion The preliminary experimental program indicated that the inclusion of polypropylene bres to HSC does not signicantly aect the compressive strength of the concrete. Compared to concrete without bres, the tensile strength was improved with all bre contents. The compressive strength of the concrete used in the experimental program just reached above the denition of HSC for the concrete with 0.3% bre content. When compared to the results of testing the concrete without bres, there was no signicant improvement in compressive strength shown by adding bres to the concrete mix. This was the same for the tensile test results. The experimental program indicated that the load carrying capacity of HSC columns was not signicantly aected through the addition of polypropylene bres. Column C6 was the only column to show an improvement in load carrying capacity of 13% when compared to Column C1. The ductility at rst spiral fracture of the columns with 0.3% bre by weight showed the most signicant improvement when compared to Column C1. The ductility was improved from adding bres to the outer cover (compared to Column C1) as in columns C4 and C7. From Table 7, it is apparent that to improve the ductility of a column it is more eective to place the bres in the outer cover as opposed to in the covercore interface. The observed failure behaviour of the columns indicated that the polypropylene bres in all cases assisted in the reduction of cracking and spalling. Columns C5, C6 and C7 failed with the spiral fracturing and the longitudinal reinforcement bending whereas the other columns did not. A bre content of 0.3% by weight is sucient to control cracking to prolong the life of a column under loading until the yield strength of the reinforcement is reached.

References
[1] Foster SJ, Attard MM. Strength and ductility of bre-reinforced high-strength concrete columns. J Struct Eng 2001;127(1):2834. [2] Foster SJ. On behaviour of high strength concrete columns: Cover spalling, steel bres, and ductility. ACI Struct J 2001;98(4):5839. [3] Sarker PK. Fibre reinforced high strength concrete columns. In: Conference proceedings: adding value through innovation, September 2001. Perth: Concrete Institute of Australia; 2001. p. 3742. [4] Foster SJ, Attard MM. Experimental tests on eccentrically loaded high-strength concrete columns. ACI Struct J 1997;94(3):295303. [5] Pessiki S, Pieroni A. Axial load behaviour of large-scale spirallyreinforced high-strength concrete columns. ACI Struct J 1997;94(3):30413. [6] Razvi SR, Saatcioglu M. Circular high-strength concrete columns under concentric compression. ACI Struct J 1999;96(5):81725. [7] Foster SJ, Liu J, Sheikh S. Cover spalling in HSC columns loaded in concentric compression. J Struct Eng 1998;124(12):14317. [8] Liu J, Foster SJ, Attard MM. Strength of tied high-strength concrete columns loaded in concentric compression. ACI Struct J 2000;97(1):14956. [9] Adepegba D, Regan P. Performance of steel bre reinforced concrete in axially loaded short columns. Int J Cement Compos Lightweight Concrete (Harlow) 1981;3(4): 2559. [10] Ganesan N, Murthy J. Strength and behaviour of conned steel bre reinforced concrete columns. ACI Mater J 1990;87(2):2217. [11] Campione G. The eects of bres on the connement models for concrete columns. Can J Civil Eng 2002;29(5):74250.