Behavior of polyester FRP tube encased recycled aggregate concrete with recycled clay brick aggregate: Size and slenderness ratio effects

Construction and Building Materials 154 (2017) 123–136

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Behavior of polyester FRP tube encased recycled aggregate concrete with recycled clay brick aggregate: Size and slenderness ratio effects Liang Huang a,⇑, Liuxin Chen a, Libo Yan b,c,⇑, Bohumil Kasal b,c, Yuxin Jiang a, Changyi Liu a a

College of Civil Engineering, Hunan University, Changsha 410082, China Centre for Light and Environmentally-Friendly Structures, Fraunhofer Wilhelm-Klauditz-Institut WKI, Bienroder Weg 54E, Braunschweig 38108, Germany c Department of Organic and Wood-Based Construction Materials, Technical University of Braunschweig, Hopfengarten 20, 38102 Braunschweig, Germany b

h i g h l i g h t s
PFRP tube significantly increases the ductility of RAC-RCBA cylinders.
Size and slenderness ratio affect the strength at the transitional point ðf ct Þ.
Wall effect existed in the small sized specimens.
Size-dependent model was proposed and fit well with the experimental results.

a r t i c l e

i n f o

Article history: Received 31 March 2017 Received in revised form 26 July 2017 Accepted 27 July 2017

Keywords: Recycled aggregate concrete (RAC) Recycled clay brick aggregate (RCBA) Polyester fiber reinforced polymer (PFRP) Confinement Compressive behavior Size effect Slenderness ratio effect

a b s t r a c t Compared with normal aggregate concrete, recycled aggregate concrete (RAC) containing recycled clay brick coarse aggregates (termed as RAC-RCBA) shows very lower compressive strength and larger variation in compressive strength, which hinder the application of RAC-RCBA as structural concrete. This study used polyester FRP (PFRP) as confining material of RAC-RCBA cylinders to improve the strength of RACRCBA. The axial compressive behavior of 42 PFRP tube encased RAC-RCBA cylinders (termed as PFRP confined RAC-RCBA) and 24 unconfined RAC-RCBA cylinders were investigated. Compared with conventional glass/carbon FRP (G/CFRP) materials, the main advantages to use PFRP are its much lower material cost and much larger tensile strain capacity. In this study, the experimental parameters considered were: (1) strength, (2) size, and (3) slenderness ratio of RAC-RCBA cylindrical specimens. Statistical analysis was also conducted to investigate the size effect and the slenderness ratio effect. The experimental results indicated that compared with conventional G/CFRP composites, the PFRP had a significantly lower tensile strength and modulus, but a much larger tensile strain at failure. The PFRP tube enhanced the ductility of the RAC-RCBA cylinders remarkably, while the enhancement on the compressive strength of RAC-RCBA by PFRP was not so pronounced as that by carbon/CFRP tubes. It was also found that the size and the slenderness ratio influenced the compressive strength of the RAC cylinders at the transitional point ðf ct Þ remarkably, and the f ct decreased with the increase in size and slenderness ratio for middle, large and tall sized cylinders. In addition, a size-dependent model for f ct was proposed and the predictions fitted well with the experimental results, the applicability of the proposed model for other weakly confined RAC-RCBA was also verified through the comparison with the experimental results of flax FRP tube confined RAC-RCBA. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction In each year, urbanization generates a huge amount of construction and demolition wastes (CDWs) in the world. Disposal of

⇑ Corresponding authors at: Department of Organic and Wood-Based Construction Materials, Technical University of Braunschweig, Hopfengarten 20, 38102 Braunschweig, Germany (L. Yan). E-mail addresses: [email protected] (L. Huang), l.yan@tu-braunsch weig.de, [email protected] (L. Yan). http://dx.doi.org/10.1016/j.conbuildmat.2017.07.197 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.

those wastes leads to an occupation and the waste of a large amount of farm lands and also causes environmental pollution issue [1]. Generally, the produce of construction materials will consume a large amount of natural resources that cannot be regenerated in a short time, thus it is a great waste if CDWs are only used for landfill. One of the most effective ways to solve CDWs disposal issue is to reuse them as recycled aggregates (RAs) of concrete, that is, to produce recycled aggregate concrete (RAC) where partial or all natural aggregates are replaced by RAs [2]. Due to a considerable amount of clay brick wastes generated in many countries in

124

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

the last decades, the recycled clay brick aggregate (RCBA) probably became one of the most widely used RA types for RAC [3–5]. To date, extensive research has been carried out on RAC because of the superiority of RAC in environmental protection and cost saving [6–10]. Particularly, a large number of studies have focused on RAC containing RCBA (termed as RAC-RCBA) [11–13]. However, owing to the RCBAs with relatively low strength and high water absorption, the strength of the produced RAC-RCBA was not satisfactory as that of normal aggregate concrete (NAC) and the compressive properties of the RAC-RCBA also showed great variation [12,13]. Consequently, the application of RAC-RCBA was mainly restricted for non-structural construction such as landfill, rather than for structural concrete application [14,15]. Fiber reinforced polymer (FRP) composite, as one of the most effective confining materials for concrete, has been widely investigated to improve the strength and ductility of concrete due to the high strength, light-weight and corrosion resistance characters of FRP materials such as glass and carbon [16–27]. In literature, a popular form of FRP confined concrete is the concrete filled FRP tube (termed as CFFT). The effects of different experimental parameters on the compressive behavior of CFFT have been extensively studied [28–37], such as strength of unconfined concrete [28,29], confinement manner [30], slenderness ratio [32], type of fiber [33], overlap configuration [34], orientation of fiber [35], end conditions [35], and section shape [36]. Recently, there has been a trend to use FRP to RAC in order to improve the mechanical property of RAC, e.g. Islam et al. [38] and Xiao et al. [39] studied the compressive behavior of CFRP and GFRP confined RAC, respectively. Their studies indicated that the confinement provided by CFRP and GFRP enhanced the strength and ductility of RACs significantly. Xie and Ozbakkaloglu [40] stated that the axial compressive stress-strain curves of CFRP confined RAC were highly affected by the replacement ratio of RAs (i.e. 50% and 100% were used), the stress-strain curves of CFRP confined RAC did not exhibit the typical bilinear behavior like the FRP confined NAC. On the other hand, Chen et al. [41] concluded that the compressive stress-strain curves of CFRP-confined RAC with different replacement ratios of RAs (i.e. 0%, 25%, 50%, 75% and 100%) were similar to that of CFRP-confined NAC and these curves can be reasonably predicted by existing stress-strain models developed for FRPconfined NAC. These limited studies mentioned here showed that the findings from different authors on the compressive behavior of G/CFRP confined RAC are conflicting. Thus, the compressive behavior of FRP confined RAC should be better understood. In practice, a wider application of CFRP and GFRP materials in civil engineering is limited by their high initial material prices. Against this background, nowadays researchers are trying to use new confining materials to replace conventional CFRP and GFRP for concrete to gain environmental and economic benefits. For example, Yan et al. [23–27] carried out a series of studies on plant-based natural flax FRP (FFRP) tube encased normal aggregate concrete and coir fiber reinforced concrete (CFRC). These researches indicated that the plant-based natural FFRP tube enhanced the bearing capacity and ductility of concrete remarkably, and also showed comparable confinement performance to those of concrete confined by GFRP and CFRP. Bai et al. [42] investigated polyethylene terephthalate (PET)-FRP and polyethylene naphthalene (PEN)-FRP confined normal concrete, the results indicated that the ductility of the confined concrete was improved remarkably due to the large rupture strain of the PET-FRP and PEN-FRP materials. Polyester is a synthetic polymer which has a large market share (i.e. 18%, ranges third place after polyethylene of 33.5% and polypropylene of 19.5%) of all plastic materials produced. The polyester has promising engineering application prospects due to its lower price (i.e. the polyester fabric is less than 1 US dollar per square meter, which is much lower than that of glass or carbon fabrics with the same size) and

the larger rupture strain (i.e. can be up to 20%) compared with synthetic carbon and glass, and plant-based natural flax fiber [43,68,69]. Although the advantages of low cost and large rupture strain capacity of polyester FRP (PFRP), it should also be pointed out that the tensile strength of PFRP is normally significantly lower compared with those of glass or carbon FRP. Whether the low strength PFRP as confining material could provide comparable confinement performance to those of glass and carbon FRP on RAC, is unknown. This is a subject which needs to be clarified. In literature, research on the application of PFRP in civil engineering is rare. Huang et al. [44] studied on the compressive behavior of PFRP confined NAC and concluded that the ductility of the concrete increased significantly and the compressive strength of concrete was also improved to some extent. Another study investigated the feasibility of PFRP plates as external strengthening materials of RC beams [45]. This study concluded that the PFRP strengthening reinforcement enhanced the ductility, deflection and energy absorption of normal RC beams remarkably. Therefore, using PFRP as confining material of RAC-RCBA may facilitate the use of RACRCBA as structural concrete, since PFRP as confinement device has potential to enhance the strength of RACs to the level of NAC. Therefore, investigating on the feasibility of PFRP tube as confining material of RAC-RCBA has potential industrial significance. Previous studies conducted by Baza˘nt et al. [46–47] have confirmed the existence of size effect in NAC and other quasi-brittle materials. Neville et al. [48] also stated that the compressive strength of small sized NAC specimens can be affected due to the restriction by the loading plates of compression testing machine, especially for specimens with a low slenderness ratio, and the restriction caused the compressive results obtained from smaller and shorter specimens not representative for larger and taller sized NAC specimens. As for FRP confined NAC, conflicted conclusions have been drawn from different researchers on its size effect. De Lorenzis et al. [49], Carey et al. [50] and Gu et al. [51] pointed out that no obvious size effect was observed in FRP confined NAC, especially in NAC confined with relatively high level confinement and large specimen sizes. However, the results from Silva et al. [52], Masia et al. [53] and Tong et al. [54] demonstrated that the size effect clearly existed and the compressive strength of FRP confined NAC decreased as the specimen size increased. Wang and Wu [55] also investigated the size effect of aramid FRP confined NAC columns and concluded that size effect actually existed when the con0 finement ratio (i.e. f l =f co , f l refers to the confinement pressure 0 provided by the FRP tube, f co indicates the compressive strength of the unconfined concrete) was relatively low, while size effect gradually disappeared as the confinement ratio increased, the critical value of confinement ratio was 0.67 according to the computation results. Most of previous studies [38–41] on FRP confined RACs were conducted on small sized (i.e. 300 mm in height and 150 mm in diameter) specimens. How the size may affect the compressive behavior of FRP confined RAC, is unknown. Against this background, this study also investigated the axial compressive behavior of PFRPconfined RAC-RCBA considering the effects of strength, size and slenderness ratio of RAC-RCBA cylinders. Furthermore, a sizedependent strength model was proposed based on Baza˘nt’s sizeeffect law [56] (Baza˘nt’s SEL) and Wang & Wu’s model [55] to predict the compressive strength of PFRP confined RAC-RCBA at the transitional point of their axial compressive stress-strain curves.

2. Experimental section 2.1. Test matrix In this study, axial compression test was carried out on 66 cylindrical specimens including 24 unconfined RAC-RCBA and 42 PFRP-

125

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

confined RAC-RCBA cylinders, the details of specimens are summarized in Table 1. The experimental parameters used were: (1) the strength of the RAC-RCBA, (2) the size of cylinders (keeping height/diameter ratio equal to 2), and (3) the slenderness ratio of cylinders (keeping cylinder diameter equal to 150 mm). In Table 1, qfrp refers to the fiber volume content of the specimens; tfrp and nfrp refers to the thickness and the number of polymer fiber layers of the PFRP tubes, respectively; f co is the compressive strength of the unconfined RAC-RCBA cylinder obtained from middle sized (i.e. 150 mm in diameter and 300 mm in height, which is the standard size used in most concrete design codes) specimens, in current study the f co of middle sized specimen is also defined as the standard compressive strength of RAC-RCBA cylinder; d and h refers to the diameter and height of the cylinders, respectively, h=d is the slenderness ratio of the RAC-RCBA specimens. All specimens were identified by three sets of characters as follows: the first set indicated the standard strength of RAC-RCBA (i.e., C2 and C3 indicated that the standard compressive strength was 25.19 MPa and 33.17 MPa for RAC-RCBA obtained from their middle sized specimens, respectively); and the second set of characters referred to the number of PFRP layers (i.e.,P0 referred to 0 layer of PFRP, that is, the specimens were unconfined, and P2, P4, P6, P8, P12 referred to PFRP tube with 2, 4, 6, 8, 12 layers of PFRP, respectively); the last set of character denoted the geometric dimensions of RAC-RCBA cylinder (i.e., S1, S2, M, L1, L2, T1, T2 indicated that the d  h of the specimens were 50  100, 100  200, 150  300, 200  400, 300  600, 150  450, 150  600 in mm, respectively). For PFRP-confined RAC-RCBA cylinders, the number of PFRP layers increased proportionally as the concrete cylinder diameter increase to keep the fiber volume content constantly, that is, to keep the same level of lateral confinement pressure provided by the PFRP tube, thus the observed values were only relevant to the size or slenderness ratio of specimens with the same standard strength. 2.2. Fabrication of specimens For unconfined RAC-RCBA cylinders, PVC tubes with designated dimensions were used as moulds for concrete casting. For PFRPconfined RAC-RCBA cylinders, Perspex tubes (i.e., a type of plastic tube with smooth surface for easily demoulding of the PFRP tubes)

were prepared as moulds to fabricate PFRP tubes, a thin layer of plastic film brushed with oil was covered on the Perspex tube for easy separation of the PFRP tube from the moulds when the PFRP tube was consolidated. Polyester fiber sheets were firstly cut into designated widths and lengths and then were saturated with epoxy resin using brushes, the resin-impregnated polyester fiber sheets were wrapped onto the Perspex tube mould. The air bubble and additional epoxy resin were squeezed out when the PFRP sheets were rolled onto the mould. The PFRP tubes were demoulded from the Perspex moulds after 24-h curing, and then the tubes were dried and cured at room temperature for seven days before concrete casting. The PFRP tubes for C3P4S2 category as an example are illustrated in Fig. 1. The natural coarse aggregates and the RAs were evenly stirred respectively to keep the water content of each type of aggregate uniformly before they were put into mixer pan. After fully mixed, the RAC-RCBA was poured into the PVC tubes and the PFRP tubes and compacted by a vibrator. After pouring, all specimens were covered by soaking wet cloth and watered three times per day for 28 days. Two narrow PFRP strips were wrapped around both ends of the specimens to avoid premature failure of the specimens. 2.3. Properties of raw materials 2.3.1. RAs and RAC-RCBA The mixture proportions of the RAC-RCBA used in this experiment are listed in Table 2. The replacement ratio of the RCBAs for the natural coarse aggregates was determined to be 70%, as previous trial tests by the authors shown that when the RAC-RCBA with 70% replacement ratio of RCBAs was used, a great balance of high replacement ratio use of RCBAs and high compressive strength of the RAC-RCBA can be achieved [3,63]. For the 70% of the RCBAs used, they contained about 60% of clay brick coarse aggregates and 40% of recycled concrete and mortar coarse aggregates by mass content. As shown in Fig. 2, the RAs were collected from Jinke Resource Recycling co. LTD in Henan Province, China. The RAs were collected from the demolished blocks which have been fully crushed, sieved and washed at Jinke’s Production line. The weak adhesion between bricks and the mortar has been destroyed in the process, which resulted in RAs containing very low adhered mortar. Tests were conducted to find the properties

Table 1 Characteristics of the tested cylindrical specimens.

a

Specimen

Number of Specimens

a

C2P0S1 C2P0S2 C2P0M C2P0L1 C2P0L2 C2P0T1 C2P0T2 C2P2S1 C2P4S2 C2P6M C2P8L1 C2P12L2 C2P6T1 C2P6T2 C3P0M C3P2S1 C3P4S2 C3P6M C3P8L1 C3P12L2 C3P6T1 C3P6T2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

0 0 0 0 0 0 0 0.0656 0.0656 0.0656 0.0656 0.0656 0.0656 0.0656 0 0.0656 0.0656 0.0656 0.0656 0.0656 0.0656 0.0656

qfrp

t frp (mm)

nfrp

f co (MPa)

d (mm)

h (mm)

h/d

0 0 0 0 0 0 0 1.72 2.89 4.25 5.12 7.06 4.25 4.25 0 1.72 2.89 4.25 5.12 7.06 4.25 4.25

0 0 0 0 0 0 0 2 4 6 8 12 6 6 0 2 4 6 8 12 6 6

25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 25.19 33.17 33.17 33.17 33.17 33.17 33.17 33.17 33.17

50 100 150 200 300 150 150 50 100 150 200 300 150 150 150 50 100 150 200 300 150 150

100 200 300 400 600 450 600 100 200 300 400 600 450 600 300 100 200 300 400 600 450 600

2 2 2 2 2 3 4 2 2 2 2 2 3 4 2 2 2 2 2 2 3 4

V frp pdnfrp t np1 h 4n t qfrp ¼ V cylinder ¼ ¼ frpd np1 , t np1 refers to the thickness of single PFRP layer which contains no epoxy resin, for this test, tnp1 = 0.41 mm. pðd=2Þ2 h

126

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

and the displacement rate was proportional to the cylinder height (0.07 mm/min, 0.13 mm/min, 0.2 mm/min, 0.27 mm/min, 0.3 mm/ min, 0.4 mm/min for cylinders with the height of 100 mm, 200 mm, 300 mm, 400 mm, 450 mm, 600 mm, respectively), to keep strain rate constantly as different strain rates may have sight influence on the compressive behavior of the cylinders. As illustrated in Fig. 4, three axial strain gauges (named by SG3, SG5, SG7, respectively) and three lateral strain gauges (named by SG4, SG6, SG8, respectively) were installed on the mid height of the cylinder surface, another two axial strain gauges (named by SG1, SG2, respectively) were installed at both ends of the cylinder surface. The applied load and vertical deformation were recorded by MTS system of the compression testing machine. The axial strain, lateral strain, applied load and vertical deformation were measured and recorded simultaneously. Fig. 1. The PFRP tubes for C3P4S2 category with tube inner diameter of 100 mm and length of 200 mm.

3. Results and discussion of the aggregates, which are summarized in Table 3. A crushing test on the RCBA indicated that the crushing index of the RCBAs was 50.9% (i.e. crushing index was defined as m1 =m0 , m0 is the gross mass of aggregates for crushing, m1 is the mass of aggregates with diameter smaller than 2.36 mm after crushing). Normal Portland cement with strength of 42.5 MPa was used for producing the RAC-RCBA. It is worth mentioning here that the recycled brick aggregates possessed significant residual cementing property, which could be attributed to partially hydrated cementitious fractions existed in RCBAs, as pointed out by Amin et al. [67]. A high cement content use for RAC-RCBA could lead to severe shrinkage cracking in the concrete, which should be paid attention to during the mix design of RAC-RCBA. 2.3.2. PFRP composites According to ASTM D3039-M08 [57], flat coupon test was conducted to determine the tensile strength, strain and modulus of the PFRP laminates, the results are displayed in Table 4. Before the test, aluminum bars of 40 mm in length and 20 mm in width were glued to the ends of the coupons to avoid premature failure of the PFRP coupon ends. More details of the flat coupons are shown in Fig. 3. The resin used in this experiment has a tensile strength of 55 MPa and tensile elasticity modulus of 3.5 GPa. As illustrated in Table 4, the tensile strength and modulus of PFPR are significantly lower than those of glass or carbon FRP composites (i.e. tensile strength of 967 MPa and elastic modulus of 60.8 GPa for GFRP, and tensile strength of 3200 MPa and elastic modulus of 213 GPa for CFRP, respectively, used in Gao et al. [63], where GFRP and CFRP tube confined RAC-RCBA were studied). However, the tensile strain of PFRP is significantly larger compared with that of glass or carbon FRP (i.e. 1.6% for GFRP and 1.5% for CFRP, respectively, according to Gao et al. [63]). 2.4. Test Instrumentation and test procedures All cylinders were tested under monotonic axial compression provided by a compression testing machine (MTS SANS YAW6506, hydraulic, Shenzhen). The loading process was executed by a displacement-control model until the failure of the specimens

3.1. Observed failure modes For unconfined RAC-RCBA cylinders, cracks emerged around the surface of the cylinders after the load was up to 60% of its peak value. The load decreased rapidly after reached its peak value and the cylinders finally failed in a brittle manner with several main vertical cracks, which was similar to that of NAC cylinders. For PFRP-confined RAC-RCBA cylinders, snapping sounds rose slowly and discontinuously during the test, and some local wrinkles emerged on the surface of the PFRP tubes. These wrinkles were attributed to the low stiffness of the PFRP tubes. At failure, the PFRP-confined RAC-RCBA cylinders failed in an explosive manner due to the rupture of PFRP tubes and the RAC-RCBA core failed in a cone type. The failure modes of the unconfined RAC-RCBA cylinders and the PFRP-confined RAC-RCBA cylinders are illustrated in Figs. 5–7, respectively. For the unconfined RAC-RCBA cylinders, no obvious differences of the failure modes were observed among cylinders with different size and slenderness ratio, while for PFRPconfined RAC-RCBA cylinders, specimens with slenderness ratio (h/ d) of 2 displayed a continuous rupture (from top to bottom) of the PFRP tubes, and specimens with slenderness ratio (h/d) of 3 and 4 exhibited localized segmented rupture of the PFRP tubes. This behavior was in agreement with the experimental phenomena reported by Vincent and Ozbakkaloglu [32]. 3.2. Compressive stress-strain behavior 3.2.1. Axial stress-strain curves Axial compressive stress-strain curves for all the specimens are demonstrated in Fig. 8. For unconfined cylinders, the axial stressstrain curves could be summarized as one type, as illustrated in Fig. 9. The figure shows an initial ascending branch to the peak load and followed by a small descending branch. The curves can be characterized by a transitional point (TP) which locates at the peak of the curves. For the tube confined cylinders, all the axial stressstrain curves can be summarized as another type, as shown in Fig. 10. The figure includes two stages: the first stage is an initial ascending linear stage which is up to the peak load while the second one is a descending nonlinear stage showing a much larger

Table 2 RAC-RCBA mixture proportions. Strength of RAC

Natural coarse aggregate (kg/m3)

Recycled coarse aggregate (kg/m3)

Natural fine aggregate (kg/m3)

Water (kg/ m3)

Cement (kg/ m3)

Water/ Cement

Replacement ratio of RAC

C2 C3

360.13 322.53

840.31 752.58

600.22 537.56

329.11 329.11

470.22 658.22

0.70 0.50

70% 70%

127

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

curves in this study indicates that the rAC-CBAC cylinders were weakly-confined by the PFRP tubes due to the much lower stiffness and tensile strength of this PFRP material. Besides, the large rupture strain of PFRP ensured that the RAC-RCBA specimen was not failure after its compressive strength reached the peak value, thus a descending branch emerged. However, it is clear that the ductility of the RAC-RCBA cylinders increased significantly due to the tube confinement, more details are discussed in Section 3.3. Fig. 8 also shows that the RAC-RCBA size and slenderness ratio have no distinct influence on the general patterns of axial compressive stress-strain curves; similar conclusion has also been drawn by Ozbakkaloglu and Vincent [31].

Fig. 2. Recycled coarse aggregates.

strain hardening compared with that of the unconfined concrete. The two stages are connected by a transition zone which reflects to the activation of the PFRP tube confinement. Different from that of the unconfined cylinders, two key points are used to characterize the curves for the tube confined cylinders, i.e. the transitional point (TP) locates at the peak of the transition zone and the ultimate point (UP) locates at the end of the curves. Unlike conventional CFRP and GFRP confined NAC typically showed an ascending second linear stage in their axial compressive stressstrain curves where this ascending linear stage was highly dependent on the tensile strength and stiffness of the confining FRP, the descending second stage in the axial compressive stress-strain

3.2.2. Axial-lateral strain responses The axial-lateral strain relationship, which displayed the dilation properties of FRP confined concrete, has been studied widely. In literature, several axial-lateral strain models of FRP confined NAC were proposed based on experimental results [58–62]. Among these models, the equations proposed by Teng et al. [61] and Lim and Ozbakkaloglu [62] were the two models which have been widely used. Thus, the predictions based on Teng et al.’s model and Lim and Ozbakkaloglu’s model were used for the comparison with the experimental axial-lateral strain curves for specimens C2P6M and C3P6M, the comparison is illustrated in Fig. 11. As can be seen from the figure, the tested axial-lateral strain curves for C2P6M firstly showed a linear branch, followed by a transitional branch. The initial linear branch reflected the dilation properties of the unconfined RAC-RCBA, i.e. the initial Poisson ratio of the unconfined RAC-RCBA was equal to 0.3. The transitional part corresponded to the activeness of PFRP tube confinement and after that, a nonlinear second branch with a rising slope was developed. Fig. 11 shows that Teng et al.’s model [61] and Lim and Ozbakkalo-

Table 3 Properties of aggregates. Aggregate type

Composition

Particle size (mm)

Moisture content

Natural coarse aggregate Recycled coarse aggregate

macadam 60% of clay brick aggregates, 40% of recycled concrete and mortar aggregates river sand

5–15 5–15

0.91% 8.09%

about 0.35–0.5

5.59%

Natural fine aggregate

Table 4 The results of PFRP flat coupon test. Number of PFRP layers

Number of specimens

Thickness (mm)

Tensile stress (MPa)

Tensile strain (%)

Elastic modulus (GPa)

2 4 6 8 12

6 6 6 6 6

1.72 2.89 4.25 5.12 7.06

31.52 37.51 40.81 43.48 41.65

8.50 11.61 14.87 16.04 17.66

0.89 0.92 0.96 0.99 0.84

Fig. 3. Configuration of a PFRP flat coupon (SG denotes for strain gauge).

128

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

Fig. 4. Test instrumentation.

Fig. 5. Typical failure mode of unconfined RAC-RCBA cylinders.

Fig. 7. Typical failure mode of PFRP-confined RAC-RCBA cylinder with h/d of 3 or 4.

considered in current study, although Teng et al.’s model [61] and Lim and Ozbakkaloglu’s model were proposed for conventional glass and carbon FRP confined NAC. Based on the equations expressed by Teng et al.’s model, a modified model was proposed to describe the axial-lateral strain responses of PFRP-confined RAC-RCBA specimens; the expression of the modified model is given below:

ec rl ¼ 0:8 1 þ 8 0 eco f co (

Fig. 6. Typical failure mode of PFRP-confined RAC-RCBA cylinder with h/d of 2.

glu’s model [62] provide a general agreement with the tendency of the experimental results of the PFRP confined RAC-RCBA cylinders



rl ¼ 

1 þ 0:2

Efrp tel R

!

 1:5 el

eco

  ) el  exp 5

eco

ð1Þ

ð2Þ

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

Fig. 8. Axial stress-strain curves for all specimens.

129

130

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

more, verification for the modified model was carried out on C3P6M specimens. As shown in Fig. 11(b), the predictions based on the proposed model also fit the experimental data of C3P6M specimens well. 3.3. Compressive test results

Fig. 9. Type of axial stress-strain curves for unconfined cylinders.

Fig. 10. Type of axial stress-strain curves for confined cylinders.

where rl and ec are the corresponding confining pressure in MPa 0 and axial strain at a given lateral strain el , respectively; f co and eco are corresponding strength in MPa and strain for unconfined concrete, respectively; Efrp and t are elasticity modulus in N=mm2 and thickness in mm of FRP tubes, respectively; R is the radius of the cylinders. Fig. 11(a) shows that the predictions based on the modified model fits the experimental data of C2P6M specimens quite well, especially has a great agreement with the increasing slope of the second nonlinear branch in the axial-lateral strain curves. Further-

The compression test results are summarized in Table 5. The cylinders of C2P0M and C3P0M categories were taken as standard specimens in the whole test matrix because they were middle sized specimens (i.e. 150 mm in diameter and 300 mm in height) which has been widely applied in various concrete design codes, and their compressive strength and corresponding axial strain were defined as f co and eco for the simplification of the following analysis. In addition, f ct and ect are compressive strength and corresponding axial strain at the transitional point (TP); f cu and ecu are compressive strength and corresponding axial strain at the ultimate point (UP). It is worth mentioning that for unconfined speci0 mens, f ct and ect represent the same meaning as f co and e0co used by other authors. The increase in compressive strength for C2P0M and C3P0M specimens was 18% and 16% due to the PFRP confinement, respectively, which were much lower than that of CFRP or GFRP confined RAC-RCBA (i.e. ranged from 101% to 347% according to Gao et al. [63] for CFRP and GFRP confined RAC-RCBA), this was attributed to the significantly lower tensile strength of the PFRP. The results reported here also coincided with the findings reported by Ozbakkaloglu and Vincent [31], which indicated that an increase in concrete compressive strength led to an overall decrease in the strength enhancement ratio for CFFT. Besides, a ductility index, l, was introduced here to describe the enhancement in ductility of RAC-RCBA due to PFRP confinement, and the value of l was equal to the ratio of ultimate axial strain of confined concrete ðecu Þ to that of axial strain at peak strength of unconfined concrete (ect or e0co used by other authors). It is known from Table 5 that the values of l ranged from 11.86 to 22.01, which were much larger than the corresponding values for CFRP and GFRP confined RAC-RCBA concluded in Gao et al. [63] (i.e. the value of ductility index ranged from 5.04 to 8.14 for GFRP and CFRP confined RACRCBA). It could also be found that l for taller specimens (C2P6T1 and C2P6T2) was smaller than that for medium specimens (C2P6M), this may be attributed to the restriction by the loading plates. Similar findings were also reported by Vincent and Ozbakkaloglu [32]. The data here indicated that the PFRP tube confinement resulted in lower increase in compressive strength but larger enhancement in ductility for RAC-RCBA cylinders, compared with those confined by glass or carbon FRP tubes. Figs. 12–15 gives the bar graphs of the tested results listed in Table 5, where the effects of size and slenderness ratio can be

Fig. 11. Comparison of model predictions with experimental axial-lateral strain curves.

131

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136 Table 5 Summaries of test results. Specimen

f co (MPa)

eco (102)

f ct (MPa)

ect (102)

f cu (MPa)

ecu (102)

f ct f co

f cu f co

ect eco

ecu eco

l

C2P0S1 C2P0S2 C2P0M C2P0L1 C2P0L2 C2P0T1 C2P0T2

25.19

0.199

16.91 22.06 25.19 24.69 23.06 24.07 23.10

0.106 0.143 0.199 0.199 0.183 0.184 0.172

– – – – – – –

– – – – – – –

0.67 0.88 1.00 0.98 0.92 0.96 0.92

– – – – – – –

0.53 0.72 1.00 1.00 0.92 0.92 0.87

– – – – – – –

– – – – – – –

C2P2S1 C2P4S2 C2P6M C2P8L1 C2P12L2 C2P6T1 C2P6T2

25.19

0.199

25.60 28.85 29.65 29.77 26.67 30.00 27.20

0.255 0.283 0.269 0.259 0.267 0.262 0.256

22.69 21.69 21.04 21.09 19.94 21.01 18.13

1.759 3.148 3.631 2.361 2.239 2.507 2.539

1.02 1.15 1.18 1.18 1.06 1.19 1.08

0.90 0.86 0.84 0.84 0.79 0.83 0.72

1.28 1.42 1.35 1.30 1.34 1.32 1.29

8.84 15.82 18.25 11.87 11.25 12.60 12.76

16.59 22.01 18.25 11.86 12.23 13.63 14.76

C3P0M C3P2S1 C3P4S2 C3P6M C3P8L1 C3P12L2 C3P6T1 C3P6T2

33.17

0.180

33.17 33.75 37.95 38.52 36.96 34.43 37.88 35.26

0.180 0.216 0.240 0.261 0.267 0.258 0.276 0.271

– 25.58 22.82 21.86 21.48 20.88 24.00 21.90

– 1.074 2.902 2.424 2.394 2.214 2.421 1.602

1.00 1.02 1.14 1.16 1.11 1.04 1.14 1.06

– 0.77 0.69 0.66 0.65 0.63 0.72 0.66

1.00 1.20 1.33 1.45 1.48 1.43 1.53 1.50

– 5.96 16.10 13.45 13.28 12.28 13.44 8.89

– – – 13.45 – – – –

reflected obviously. In Figs. 12–15, C2-U represents the unconfined concrete group with standard compressive strength of 25.19 MPa (measured by the medium (or standard) sized cylinder, i.e. 150 mm in diameter and 300 mm in height); C2-P refers to the PFRP confined concrete group with the standard unconfined compressive strength of 25.19 MPa; C3-P represents the PFRP confined concrete group with the standard unconfined compressive strength of 33.17 MPa. S1, S2, M, L1 and L2 refer to the specimen category with the dimension of 50  100, 100  200, 150  300, 200  400, 300  600 in mm, respectively. T1 and T2 refer to the specimen category with dimension of 150  450 and 150  600 in mm, respectively. It is observed from Fig. 12(a) that, apart from the categories S1 and S2, the compressive strength decreases with an increase in size of the RAC-RCBA cylinders. This might be attributed to the ‘‘wall effect” [48] existed in the small sized concrete cylinders. If the maximum size of aggregate is large in relation to the size of the mould for concrete, the compaction of concrete and the uniformity of distribution of the large particles of aggregate are affected. This is known as the ‘‘wall effect” because the wall influences the packing of the concrete; the quantity of mortar required to fill the space between the particles of the coarse aggregate and the wall is greater than that necessary in the interior of the mass, and therefore greater than the quantity of mortar available in a wellproportioned mix, as explained by Neville [48]. The wall effect is more pronounced, the larger the surface/volume ratio of the concrete specimen is. Thus, to minimize the wall effect, various

concrete standards specify the minimum size of the test specimen in relation to the maximum size of aggregate, i.e. British Standard BS 1881-103: allow the use of 100 mm cubes and 100 by 200 mm cylinders with aggregates whose maximum size is up to 20 mm; 150 mm cubes and 150 by 300 mm cylinder can be used with aggregate up to 40 mm in size. The requirement of ASTM C19207 is that the diameter of the test cylinder or the minimum dimension of a prism be at least 4 times the nominal maximum size of aggregate [48]. It should be pointed out here that for the coarse aggregate used in this study, the maximum size is 15 mm. The diameter of the small sized concrete cylinder was 50 mm. The ‘‘wall effect”, as explained by Kreijger [64], corresponds that the volume fraction of aggregates adjacent to the formwork is lower than that inside of the concrete, this results in the existing of surface layer of concrete which has lower quality than that of the concrete in the inner part of the cylinder. The surface layer, as recommended by Kreijger [64], has three sublayers, the cement sublayer, the mortar sublayer and the concrete sublayer. Mamilan [65] further pointed out that the concrete at surface layer has a decrease in elasticity modulus and increase in porosity compared with the inner part of concrete cylinder, and the depth of surface layer is about 30–40 mm for normal aggregate concrete. Therefore, the concrete at surface layer has lower strength capacity than the inner part of concrete, for large sized specimens, the lower strength capacity of surface layer has no significant influence on the compressive strength of the cylinders, while for small sized specimens; the influence of the surface layer is remarkable and reduces the

Fig. 12. Comparison on f ct for specimens with different size and slenderness ratio.

132

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

Fig. 13. Comparison on f cu for specimens with different size and slenderness ratio.

Fig. 14. Comparison on

ect for specimens with different size and slenderness ratio.

Fig. 15. Comparison on

ecu for specimens with different size and slenderness ratio.

compressive strength of the concrete specimen greatly. In Fig. 12 (b), a decrease trend is observed on f ct with the increase of slenderness ratio, and such trend is also observed on f cu in Fig. 13 as the size and slenderness ratio increase. The notable exception is C3-P group, in which f cu for M category is lower than that for T1 and T2 categories. The ecu also displays a descending trend in Fig. 15 as the size and slenderness ratio increase for medium, large and tall sized specimens. No obvious tendency is observed in Fig. 14 on ect varying with specimen size and slenderness ratio. However, the overall decrease trend on f ct , f cu and ecu observed from the bar graphs could not confirm that the size and slenderness ratio effects really exist, because the random error may hide in this experiment and affect the experimental results. Further research is still needed to better understand the size and slenderness ratio effect in a mathematical sense. 4. Statistical analysis From the last section we know that the f ct , f cu and ecu displayed descending trends as specimen size and slenderness ratio increase in some cases, however, whether these trends are affected by size

and slenderness ratio effect or random error, cannot fully confirmed. Besides, although ect has not shown a clear trend, we cannot verify that specimen size and slenderness ratio have no influence on theect . Therefore, further investigation in mathematical sense is essential. Single factor variance analysis, as a strict and effective mathematical analysis method, is used in the present study to investigate whether the influences of specimen size and slenderness exist in different categories. For n sampled data divided into r groups, each group is corresponded one level of the inspected value. The total variations of all the sampled data (termed as ST Þ could be divided into two parts: the variations from random error (termed as SE Þ and the variation from different groups (termed as SG Þ. If the level of the inspected value really affects the sampled data (i.e., there exist differences between the inspected value from different groups), the SG should be relatively larger than SE , otherwise the inspected value has no influence on the sampled data. The quantitative formula of ST , SE and SG are exhibited below:

ST ¼

ni r X X  i¼1 j¼1

X ij  X

2

ð3Þ

133

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

SE ¼

ni r X X 

X ij  X i

2

ð4Þ

Table 7 Results of single factor variance analysis.

i¼1 j¼1

SG ¼

ni r X X



Xi  X

2

ð5Þ

i¼1 j¼1

ST ¼ S E þ S G

Inspected value

Influence of size

Influence of slenderness ratio

Influence of wall effect

f ct f cu

Confirmed Questionable Questionable Questionable

Confirmed Questionable Questionable Questionable

Confirmed – – –

ect ecu

ð6Þ

where X ij is the j set of concrete value in group i of all the sampled data, X is the average value of all the sampled data, X i is the average value of the data in group i, ni is the number of sampled data in group i. The process of single factor variance analysis is illustrated in Table 6, firstly a hypothesis is given that there is no difference of the inspected value from different groups, then the variations from random error ðSE Þ and the variation from different groups ðSG Þ are calculated through the Eqs. (3)–(6). For the purpose of quantitative analysis, the F-value is proposed to represent the relative magnitude of SG compared with SE , then a P-value is calculated to indicate the probability of the hypothesis. The equations for F-value and Pvalue are exhibited in Table 6. The confidence level is taken as 95% (d = 0.05), if P-value 6 0:05, the hypothesis should be rejected, if Pvalue P 0:95, the hypothesis should be accepted, or else the hypothesis is still questionable. The results of single factor variance analysis are exhibited in Table 7, the wall effect has also been investigated through the comparison of S1, S2 and M categories. It could be concluded from Table 7 that size and slenderness ratio actually have influences on f ct and the manifestation of these influences is that f ct decreases as the specimen size or the slenderness ratio increases for middle, large and tall sized cylinders. There is no enough evidence show that the size and the slenderness ratio have influence on f cu , ect and ecu although they display descending trends as specimen size and slenderness ratio increase in some cases. This may be attributed to the limited test data in the present study, further studies are still needed to investigate the size effect and slenderness ratio effect on f cu , ect and ecu . 5. Size-dependent model and verification The discussions in last section confirmed the existence of size effect and slenderness ratio effect on f ct . Therefore, a model that not only considers the confinement ratio but also considers the size and slenderness ratio effect should be created for the design purpose for f ct . According to size effect analysis based on the crack band model, an equation was proposed for plain concrete by Baza˘nt [56] as follows:

B f

0 t ffi rN ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ð7Þ

1 þ k0Dda

where rN is nominal stress, f t is tensile strength of concrete, D is characteristic dimension in mm, B0 and k0 are empirical constants, da is maximum aggregate size in mm. After that, an equation with stronger applicability was proposed by Kim et al. [66] based on Baza˘nt’s model, and the model, which substitutes f t and D for f co and h  bd, is not only easier to be

applied using compressive strength f co , but also is applicative to fit the results from specimens with different slenderness ratios. The equation is given below:

B f

1 co ffi rN ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ð8Þ

1 þ hbd k1 da

where B1 , k1 and b are empirical constants, and b is usually assumed to be 1. By considering B1 as the influence of the confinement attributed to AFRP jackets and k1 as magnitude of the size effect, Wang and Wu [55] applied Baza˘nt’s and Kim’s models for AFPR confined concrete in a creative way. Through fitting the result data with the modified model, Wang and Wu proposed an equation suitable to predict the f ct for AFRP confined concrete with different sizes:

f ct



f 1:2 þ 3:85 f colf f co qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ 1 þ hd 545

ð9Þ

where f lf is confinement pressure provided by AFRP jackets, here f lf has the same meaning with the f l mentioned above. However, Wang and Wu’s model was derived from the results of NAC, and also the concrete confined by AFRP material in Wang and Wu’s experiment showed a typical ascending tendency in the second linear stage at the axial compressive stress-strain curves which were classified into strongly confined specimens, thus it was not suitable to use Wang and Wu’s model directly to fit the test results from PFRP-confined RAC-RCBA. Thus, based on Wang and Wu’s method, a modified model was developed to fit the test results in the present study, as illustrated in Fig. 16. It should be pointed out here that the small sized specimens exhibited the wall effect clearly and the f ct derived from them was lower than the specimens with larger sizes, which was contrast to the size effect law of the unconfined concrete. Especially for S1 category, that caused the maximum discrepancy from the size effect law. In consideration of the rareness use of such small size of S1 in laboratory and practical engineering, the results from S1 category were not included in the fitting analysis. Therefore, the modified model to fit the test results of PFRP confined RAC-RCBA can be exhibited as follows:

f ct



f 1:05 þ 2:25 f lf f co qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffico ¼ hd 1 þ 1354

ð10Þ

The performance of the modified model is illustrated in Fig. 17. It is known from Figs. 16 and 17 that the modified model fits the experimental results quite well. Moreover, according to the Eq.

Table 6 Process of single factor variance analysis. Source of variance

Sum of squares

Degree of freedom

Mean square

F-value

P-value

Judgment of P-value

Results

Among groups Intra-group Total

SG SE ST

r1 nr n1

SG =ðr  1Þ SE =ðn  rÞ

SG =ðr1Þ SE =ðnrÞ

F 1 ½ðr1Þ;ðnrÞ ðF  v alueÞ

6 0:05 0:05  0:95 P 0:95

Rejected Questionable Accepted

134

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

[3]. In the future study, more extensive database of weakly confined RAC-RCBA will be needed for further verification of the modified model. 6. Conclusions This paper has reported the results of an experimental study on the axial compression behavior of polyester FRP tube encased recycled aggregate concrete containing recycled clay brick aggregates, the influence of size and slenderness ratio of RACs on the compressive behavior of PFRP-confined RAC-RCBA were investigated. The findings indicated that size and slenderness ratio have influence on the compressive strength of the cylinders at the transitional point ðf ct Þ, the wall effect has also been verified to exist in the small sized cylinders. These findings were in agreement with majority related studies in the existing literature on FRP confined NAC, e.g. Ref. [31,32,37]. Based on the results and analysis in this study, the following conclusions can be drawn:

Fig. 16. Fitting results for f ct .

Fig. 17. Performance of the modified model.

(10), the modified model could not only provide predictions for the f ct of the specimens with different size and slenderness ratio but also with different confinement ratios. To verify the applicability of the modified model for weakly confined RAC-RCBA, a comparison with Yan et al. [3] results (where the behavior of natural flax FRP tube confined RAC-RCBA was considered) was carried out, and the comparison is listed in Table 8. Table 8 shows that the modified model also provided a relatively accurate prediction for the experimental results of flax FRP tube confined RAC-RCBA presented in

(1) PFRP composites showed significantly lower tensile strength and modulus but much larger tensile strain at failure compared with conventional glass and carbon FRP composites. (2) The use of PFRP tube significantly increased the ductility of RAC-RCBA cylinders, the ductility index ranged from 11.86 to 22.01, which was much larger than the range of 5.04– 8.14 for synthetic CFRP and GFRP tube confined RAC-RCBA [63]. The compressive strengths of all the confined RACRCBA cylinders showed a growth to a relatively low extent due to the PFRP confinement, which were 18% and 16% for middle sized specimens, much lower than 101%–347% of GFRP and CFRP tube confined RAC-RCBA due to the much lower tensile strength of PFRP compared with those of GFRP and CFRP. (3) Size and slenderness ratio affected the strength at the transitional point ðf ct Þ of the test cylinders, and the findings indicated that f ct decreased as the size or the slenderness ratio increased for middle, large and tall sized cylinders. No enough evidence showed that the size and slenderness ratio had influence on ultimate strength ðf cu Þ, strain at the transitional point ðect Þ and ultimate strain ðecu Þ, further studies are still needed to investigate this aspect. (4) Small sized specimens (i.e. 50 mm in diameter, 100 mm in height, and 100 mm in diameter, 200 mm in height) contributed larger variations to strength at the transitional point ðf ct Þ, and this may be attributed to the wall effect existed in the small sized specimens, that is, the low strength capacity of surface layer of cylinders reduced the compressive strength of small specimens significantly.

Table 8 Comparisons of the predicted values with Yan et al.’s [3] results. Source of Data

Categories

h (mm)

d (mm)

f co (MPa)

f ct:exp (MPa)

f ct:pre (MPa)

f ct:pre f ct:exp

Mean value

Std.

Yan et al.3

RC2F0-M RC2F3-M RC2F6-M RC2F9-M RC3F0-S RC3F3-S RC3F0-M RC3F3-M RC3F6-M RC3F9-M RC3F0-L RC3F12-L

300 300 300 300 150 150 300 300 300 300 600 600

150 150 150 150 75 75 150 150 150 150 300 300

27.54

27.54 29.00 32.82 37.27 23.28 36.88 32.84 38.49 43.54 49.47 27.67 45.05

27.44 31.92 35.98 38.37 33.56 42.30 32.72 37.20 41.26 43.65 31.20 39.32

1.00 1.10 1.10 1.03 1.44 1.15 1.00 0.97 0.95 0.88 1.13 0.87

1.05

0.146

32.84

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136

(5) A size-dependent model was proposed in this paper based on Baza˘nt’s SEL [56], Kim et al. [66] and Wang and Wu’s [55] model, the model predictions fitted well with the experimental results in present study, and the applicability of the proposed model has been verified through the comparison with test data of flax FRP tube confined RAC-RCBA in Ref. [3]. Overall, current study showed that the low tensile strength PFRP composite as confining material can increase the compressive strength of RAC-RCBA with a high replacement ratio of natural aggregates (i.e. 70%) to a normal concrete strength level which made the RAC-RCBA with potential to be used as structural concrete. In addition, it cannot be ignored that the PFRP confinement can significantly enhance the ductility of RAC-RCBA, which is favorable for structural design where high demand of deformation capacity of structural concrete is required, i.e. in the design of seismic-resistant structures. A potential application of PFRP confined RAC-RCBA could be energy dissipation components for buildings. In addition, the use of cost-effective PFRP and RAC-RCBA may save construction cost and have a great significance for environmental protection and natural resource saving as construction and building materials. Future research is required to fully understand the fire performance, seismic performance and energy dissipation capacity of PFRP tube confined RAC-RCBA structural components.

Acknowledgments This study is supported by Hunan Provincial Natural Science Foundation of China (No.2015JJ1004). The writers are also grateful for the technical support from staff in Structural Laboratory of JINKE Resource Recycling Company in Xu Chang, China.

References [1] P. Tang, P. Yang, Analysis of industrialization of C&D waste disposition in China, Jiangsu Constr. 3 (2007) 57–60. [2] M.T. Uddin, H. Ariful, A.M. Abdul, Z.B. Shamim, Recycling of brick aggregate concrete as coarse aggregate, J. Mater. Civ. Eng. 27 (7) (2015) B4014005. [3] B. Yan, L. Huang, L.B. Yan, C. Gao, B. Kasal, Behavior of flax FRP tube encased recycled aggregate concrete with clay brick aggregate, Constr. Build. Mater. 136 (2017) 265–276. [4] M. Wahlstrom, J. Laine-Ylijoki, A. Maattanen, Environmental quality assurance system for use of crushed mineral demolition wastes in road constructions, Waste Manage. 20 (2) (2000) 225–232. [5] Municipality of Mashhad, Annual report of construction and demolition debris. Technical Rep. No. RTMO-85, Research Centre for Mashhad City Council, Mashhad, Iran, 2007. [6] E. Marsh, Civil infrastructure systems materials research support at the national science foundation, Cem. Concr. Compos. 25 (6) (2003) 575–583. [7] K.K. Sagoe-Crentsil, T. Brown, A.H. Taylor, Performance of concrete made with commercially produced coarse recycled concrete aggregate, Cem. Concr. Res. 31 (5) (2001) 707–712. [8] W.C. Choi, H.D. Yun, Compressive behavior of reinforced concrete columns with recycled aggregate under uniaxial loading, Eng. Struct. 41 (3) (2012) 285– 293. [9] J.Z. Xiao, Recycled concrete, China Architecture & Building Press, Beijing, China (in Chinese), 2008. [10] C.S. Poon, S.C. Kou, L. Lam, Use of recycled aggregates in molded concrete bricks and blocks, Constr. Build. Mater. 16 (5) (2002) 281–289. [11] F.M. Khalaf, Using crushed clay brick as coarse aggregate in concrete, J. Mater. Civ. Eng. 18 (4) (2006) 518–526. [12] F.M. Khalaf, A.S. Devenny, Properties of new and recycled clay brick aggregates for use in concrete, J. Mater. Civ. Eng. 17 (4) (2005) 456–464. [13] F.M. Khalaf, A.S. Devenny, Recycling of demolished masonry rubble as coarse aggregate in concrete: review, J. Mater. Civ. Eng. 16 (4) (2004) 331–340. [14] R.J. Collins, P.T. Sherwood, Use of waste and recycled materials as aggregates: Specifications and standards, Dept. of the Environment Rep, HMSO, London, 1995. [15] A.D. Buck, Recycled concrete as a source of aggregate. 74 (1976) 212–219. [16] L. Huang, C. Gao, L.B. Yan, B. Kasal, G. Ma, (2016) Reliability assessment of confinement models of carbon fiber reinforced polymer-confined concrete, J. Reinf. Plast. Compos. 35 (12) (2016) 996–1026.

135

[17] L.B. Yan, Plain concrete cylinders and beams externally strengthened with natural flax fabric reinforced epoxy composites, Mater. Struct. 49 (6) (2016) 2083–2095. [18] P. Yin, L. Huang, L. Yan, D. Zhu, Compressive behavior of concrete confined by CFRP and transverse spiral reinforcement. Part A: experimental study, Mater. Struct. 49 (3) (2016) 1001–1011. [19] L. Huang, X. Sun, L. Yan, B. Kasal, Impact behavior of concrete columns confined by GFRP tube and steel spiral reinforcement, Constr. Build. Mater. 131 (2017) 438–448. [20] L. Yan, N. Chouw, Effect of water, seawater and alkaline solution ageing on mechanical properties of flax fabric/epoxy composites used for civil engineering applications, Constr. Build. Mater. 99 (2015) 118–127. [21] L. Huang, X. Xun, L. Yan, D. Zhu, Compressive behavior of concrete confined with GFRP tubes and steel spirals, Polymers 7 (5) (2015) 851–875. [22] C. Gao, L. Huang, L. Yan, G. Ma, L. Xu, Compressive behavior of CFFT with inner steel wire mesh, Compos. Struct. 133 (2015) 322–330. [23] L. Yan, N. Chouw, Experimental study of flax FRP tube encased coir fibre reinforced concrete composite column, Constr. Build. Mater 40 (2013) 1118– 1127. [24] L. Yan, N. Chouw, K. Jayaraman, Effect of column parameters on flax FRP confined coir fibre reinforced concrete, Constr. Build. Mater 55 (2014) 213– 299. [25] L. Yan, N. Chouw, Compressive and flexural behaivor and theoretical analysis of flax fibre reinforced polymer tube encased coir fibre reinforced concrete composite, Mater. Des. 52 (2013) 801–811. [26] L. Yan, N. Chouw, Dynamic and static properties of flax fibre reinforced polymer tube confined coir fibre reinforced concrete, J. Compos. Mater 48 (13) (2014) 1595–1610. [27] L. Yan, N. Chouw, Natural FRP tube confined fibre reinforced concrete under pure axial compression: A comparison with glass/carbon FRP, Thin-Walled Struct. 82 (2014) 159–169. [28] T. Ozbakkaloglu, D.J. Oehlers, Concrete-filled square and rectangular FRP tubes under axial compression, J. Compos. Constr. 12 (4) (2008) 469–477. [29] T. Ozbakkaloglu, J.C. Lim, T. Vincent, FRP-confined concrete in circular sections: Review and assessment of stress-strain models, Eng. Struct. 49 (2013) 1068–1088. [30] T. Vincent, T. Ozbakkaloglu, Influence of concrete strength and confinement method on axial compressive behavior of FRP confined high and ultra-high strength concrete, Compos. Part B 50 (2013) 413–428. [31] T. Ozbakkaloglu, T. Vincent, Axial compressive behavior of circular high strength concrete-filled FRP tubes, J. Compos. Constr. 18 (2) (2013) 04013037. [32] T. Vincent, T. Ozbakkaloglu, Influence of slenderness on stress-strain behavior of concrete filled FRP tubes: experimental study, J. Compos. Constr. 19 (1) (2014) 04014029. [33] T. Vincent, T. Ozbakkaloglu, Influence of fibre type on behavior of HSC-filled FRP tubes under concentric compression, Appl. Mech. Mater. 438 (2014) 240– 245. [34] T. Vincent, T. Ozbakkaloglu, Influence of overlap configuration on the compressive behavior of CFRP confined normal-and high-strength concrete, Mater. Struct. (2015) 1–24. [35] T. Vincent, T. Ozbakkaloglu, Influence of fibre orientation and specimen end condition on axial compressive behavior of FRP confined concrete, Constr. Build. Mater. 47 (2015) 814–826. [36] T. Ozbakkaloglu, Behavior of square and rectangular ulta high-strength concrete-filled FRP tubes under axial compression, Compos. B 54 (2015) 97– 111. [37] J.C. Lim, T. Ozbakkaloglu, Influence of size and slenderness on compressive strain softening of confined and unconfined concrete, J. Mater. Civ. Eng. 28 (2) (2016) 06015010. [38] M.M. Islam, M.S.I. Choudhury, M. Abdulla, A.F.M.S. Amin, Confinement effect of fiber reinforced polymer wraps in circular and square concrete columns. in: 4th Annual Paper Meet and 1st Civil Engineering Congress, the Institution of Engineers (IEB), Dhaka, Bangladesh, (2011) 359–362. [39] J.Z. Xiao, Y.J. Huang, J. Yang, C. Zhang, Mechanical properties of confined recycled aggregate concrete under axial compression, Constr. Build. Mater 26 (1) (2012) 591–603. [40] T.Y. Xie, T. Ozbakkaloglu, Behavior of recycled aggregate concrete-filled basalt and carbon FRP tubes, Constr. Build. Mater. 105 (2016) 132–143. [41] G.M. Chen, Y.H. He, T. Jiang, C.J. Lin, Behavior of CFRP-confined recycled aggregate concrete under axial compression, Constr. Build. Mater. 111 (2016) 85–97. [42] Y.L. Bai, J.G. Dai, J.G. Teng, Cyclic compressive behavior of concrete confined with large rupture strain frp composites, J. Compos. Constr. 18 (1) (2014) 04013025. [43] H. Jaqin, H. Nakai, T. Ueda, Y. Sato, J. Dai, Seismic retrofitting of RC piers using continuous fiber sheet with large fracturing strain, J. Struct. Eng. 51 (2001) 893–902. [44] L. Huang, X. Yang, L. Yan, K. He, H. Li, Y. Du, Experimental study of polyester fiber-reinforced polymer confined concrete cylinders, Text. Res. J. 86 (15) (2016) 1606–1615. [45] L. Huang, L.K. Zhao, L.B. Yan, Flexural performance of RC beams strengthened with polyester FRP composites, Int. J. Civ. Eng. (2017), http://dx.doi.org/ 10.1007/s40999-016-0140-0. [46] Z.P. Bazant, Size effect in tensile and compression fracture of concrete structures: Computational modeling and design, in: F.H. Wittmann (Ed.),

136

[47] [48] [49]

[50]

[51] [52]

[53] [54]

[55] [56] [57] [58]

L. Huang et al. / Construction and Building Materials 154 (2017) 123–136 Fracture Mechanics of Concrete Structures, Freiburg, Germany, Aedificatio Publishers, pp. 1905–1922. Z.P. Bazant, J. Planas, Fracture Size Effect in Concrete and Other Quasibrittle Materials, CRC Press, Boca Raton, 1998. A.M. Neville, Properties of concrete, Longman Group Ltd., Essex, U.K., 1995. L.D. Lorenzis, F. Micelli, A.L. Tegola, Influence of specimen size and resin type on the behavior of FRP-confined concrete cylinders, in: Proc., Advanced Composites for Structural Applications in Construction (ACIC-2002), Univ. of Southampton, Southampton, UK, 2002, pp. 46–53. S.A. Carey, K.A. Harries, Axial behavior and modeling of confined small-, medium-, and large-scale circular sections with carbon fiber-reinforced polymer jackets, ACI Struct. J. 102 (4) (2005) 596–604. X.L. Gu, Compressive stress-strain relationship of concrete confined by carbon fiber composite sheets, Struct. Eng. 22 (2) (2006) 50–56. M.A. Silva, C.C. Rodrigues, Size and relative stiffness effects on compressive failure of concrete columns wrapped with glass frp, J. Mater. Civ. Eng. 18 (3) (2006) 334–342. M.J. Masia, T.N. Gale, N.G. Shrive, Size effects in axially loaded square-section concrete prisms strength, Can. J. Civ. Eng. 31 (31) (2004) 1–13. G.S. Tong, Y.S. Liu, Q.L. Wu, Size effect on square section column concrete confined with basalt fiber sheets, Concrete (2009), http://dx.doi.org/10.3969/j. issn.1002-3550.2009.03.002. Y.F. Wang, H.L. Wu, Size effect of concrete short columns confined with aramid frp jackets, J. Compos.Constr. 15 (4) (2011) 535–544. Z.P. Baza˘nt, Size effect in blunt fracture: concrete, rock, metal, J. Eng. Mech 110 (4) (1984) 518–535. D. Astm, Standard test method for tensile properties of polymer matrix composite materials (2008). A. Mirmiran, A. Singhvi, G. Monti, Frp-confined concrete model, J. Compos. Constr. 3 (3) (1999) 143–150.

[59] A.Z. Fam, S.H. Rizkalla, Confinement model for axially loaded concrete confined by circular fiber-reinforced polymer tubes, ACI Struct. J. 98 (4) (2001) 451–461. [60] S.P.C. Marques, D.C.S.C. Marques, J.L. da Silva, M.A.A. Cavalcante, Model for analysis of short columns of concrete confined by fiber-reinforced polymer, J. Compos. Constr. 8 (4) (2004) 332–340. [61] J.G. Teng, Y.L. Huang, L. Lam, L.P. Ye, Theoretical model for fiber-reinforced polymer-confined concrete, J. Compos. Constr. 11 (2) (2007) 201–210. [62] C.L. Jian, T. Ozbakkaloglu, Lateral strain-to-axial strain relationship of confined concrete, J. Struct. Eng 141 (2014) 04014141. [63] C. Gao, L. Huang, L. Yan, B. Kasal, W. Li, Behavior of glass and carbon frp tube encased recycled aggregate concrete with recycled clay brick aggregate, Compos. Struct. 155 (2016) 245–254. [64] P.C. Kreijger, The skin of concrete composition and properties, Mater. Struct. 17 (4) (1984) 275–283. [65] M. Mamillan, M. Mamillan, The heterogeneity of concrete revealed by the speed of sound, Ann. Tech. Inst. Build. Public Works, nr 309–310 (1973) 35–38 (In French). [66] J.K. Kim, S.T. Yi, C.K. Park, S.H. Eo, Size effect on compressive strength of plain and spirally reinforced concrete cylinders, ACI Struct. J. 96 (1) (1999) 88–94. [67] A.F.M.S. Amin, A. Hasnat, A.H. Khan, M. Ashiquzzaman, Residual cementing property in recycled fines and coarse aggregates: Occurrence and quantification, J. Mater. Civ. Eng. 28 (4) (2016) 04015174. [68] L. Yan, N. Chouw, K. Jayaraman, Flax fibre and its composites – a review, Compos. B 56 (2014) 296–317. [69] J. Sliseris, L. Yan, B. Kasal, Numerical modelling of flax short fibre reinforced and flax fibre fabric reinforced polymer composites, Compos. B 89 (2016) 143– 154.