- Email: [email protected]
Interfacial properties and structural performance of resin-coated natural fibre rebars within cementitious matrices
Cement and Concrete Composites 87 (2018) 44e52
Contents lists available at ScienceDirect
Cement and Concrete Composites journal homepage: www.elsevier.com/locate/cemconcomp
Interfacial properties and structural performance of resin-coated natural fibre rebars within cementitious matrices Hassan Ahmad, Mizi Fan* Nanocellulose and Biocomposites Research Centre, College of Engineering, Design and Physical Sciences, Brunel University London, UB8 3PH, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 October 2016 Received in revised form 24 November 2017 Accepted 4 December 2017 Available online 6 December 2017
This paper investigates the interface and behaviour of continuous natural fibre rebar reinforced cementitious composites. Various resins were used to coat sisal fibres which were then used as reinforcement within the composites. Non-destructive tests (SEM, OM, FTIR) were employed to evaluate the interfacial bonding and interaction between the sisal, resin and cementitious matrix. The results showed that the uncoated fibre could not develop a compact interface due to the moisture absorption/desorption of the fibre, which caused low mechanical properties of the composite. The coating was able to improve the strength of the rebars and reduce the OH concentration, which improved the interfacial bonding and integrity. The flexural stress-strain relation of the composites exhibited three regions with two clear dips, reflecting the progressive stress transfer and failure of the composite constituents and their interactions. The comparable mechanical properties to those of steel rebar reinforcement demonstrate the potential of the resin-coated sisal fibre rebar for structural applications. © 2017 Published by Elsevier Ltd.
Keywords: Sisal fibre Resin coating/surface treatment Natural fibre rebar cementitious composite Fibre/matrix bond Interface Microstructural analysis
1. Introduction Various forms of fibres have been developed to overcome the limitations of concrete as a quasi-brittle material that is weak in tension. Fibres typically act to restrict crack growth within concrete. Natural fibres are traditionally classified into four categories according to their origin: stem/bast, leaf, fruit/surface and wood/ cellulose [1]. Leaf fibres provide the fundamental rigidity and strength of leaves and are extracted by means of scraping and drying crushed leaves. In comparison to stem/bast fibres, the leaf fibres are commonly tougher, rougher in texture and stronger on average. For example, the strength and stiffness of sisal fibre range from 550 to 750 MPa and 17e38 GPa compared to those of jute fibre from 300 to 800 MPa and 10e30 GPa respectively [2e8]. Sisal fibres, being the most common type of leaf fibres, are produced from various varieties of the agave plant, which is indigenous to the arid regions of Africa and South America, mainly Brazil e the biggest commercial producer of Agave sisalana amounting to an estimate of 150,000 tonnes [9]. Known to have one of the strongest natural fibres, the agave plant is characterised by its
* Corresponding author. E-mail addresses: [email protected] (H. Ahmad), [email protected]. uk (M. Fan). https://doi.org/10.1016/j.cemconcomp.2017.12.002 0958-9465/© 2017 Published by Elsevier Ltd.
leaves, which grow to a length of over one metre, and yield a long creamy white and very strong fibre. Sisal is a fast-growing plant that remains reproductive for 10e12 years and produces 180 to 240 leaves in a lifetime. In addition, having a total annual production of approximately 300,000 tonnes with a low-cost production, sisal fibre is increasingly being used in the building industry, particularly for plaster reinforcement. As sisal has been used for many years for reinforcing gypsum products, there is an emerging potential for sisal fibres being used to reinforce other composites for low-cost housing applications, namely reinforcing concrete matrices [10e13]. The characteristics and performance of the fibre-reinforced composites depend on the properties of the individual components within the composites including the fibre, the individual matrices and the interfaces between them. One must, therefore, investigate the nature of each of these components to gain a comprehensive understanding of the properties of the final composite. When considering vegetable fibre reinforced composites, the major research challenge has been trying to achieve a strong interface, especially a dense interfacial transition zone, within the composites. However, the surface chemistry of natural fibres and the negative effects of moisture diffusion between the fibre and the inorganic-bonded-matrix (e.g. mortar) on the interfacial bonding complicate this. In realising and optimising the requirements for
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
exceptional bonding mechanism, i.e. mechanical, electrostatic, interdiffusion or chemical bonding between the fibre and matrix [14], the fibre and matrix can such reach their individual strength such that the load applied is uniformly distributed within the lattice structure [15]. Moreover, for natural fibre-inorganic composites, there exists an additional major challenge: hydrophilic nature of natural fibre complicating with the essential presence of moisture for the inorganic matrix (requirement for curing). Hence, this paper is to systematically analyse and characterise the interfacial systems of the natural fibre (sisal fibre as an example) e inorganic matrix (cement-based as a candidate) composites (FRC). The findings of this research will result in a better understanding of the interface and interaction of natural fibre with the cementitious constituent, provide fundamental information for further development of the natural fibre inorganic composites, and such facilitate and promote the use of natural fibres as a structural reinforcement for cementitious composites that could be a viable alternative to steel in some applications. 2. Experimental plan 2.1. Raw materials Natural fibre sisals (Agave sisalana) were purchased online in a form of ropes comprised of 3-ply twined yarns to make up 3 mm diameter string of the desired length. It is graded as type-1 sisal, which is the finest known quality. Additionally, four types of resins (slow cure time) were bought from a commercial supplier for coating sisal fibres; epoxy, polyurethane, vinylester and polyester. The properties of the resins can be summarised in Table 1. 3 mm marine grade steel wire rope was used as a comparison. Standard CEM II Portland-limestone (~6e20%) cement (Class 32,5 R), complying with BS EN 197-1 [16] was used for making the matrix along with sharp sand bought from a commercial building material supplier.
45
fibre and resin, the effectiveness of which has been proven by some studies [17e22]. Samples of uncoated steel rebar were fitted in the same way. The mix proportion used for making the cementitious matrix used was a 1:1.5 ratio of cement to sand with water to cement (w/c) ratio of 0.4. The sand used was oven dried at 100 C for 24 h and sieved to a particle size distribution of 1 mm. The fresh mortar was fabricated using a bench-top mixer before casting into the moulds and compacting for 1 min using a vibrating table, conforming to EN 196-1 [23]. The moulds were then covered with a plastic film and water sprayed on top of the plastic film to protect and control the moisture loss. After 24 h, the cementitious composites were de-moulded and left to cure at room temperature and pressure of 22 ± 1 C and ~100 kN/m2 with relative humidity of 50% within a curing chamber till the ages of 7, 14 and 28 days. To improve the reliability of the investigation, four samples of each type of composite were tested and the mean values were used. Control variables were maintained, such as keeping the proportions of the ingredients the same and applying the same amount of resin used on the rebar (6 g of any resin applied for every 60 cm length of twined sisal rope), such that coated sisal rebars have diameters of 3.5 ± 0.2 mm when measured via a Vernier caliper (~0.3 ± 0.02 vol% of fibre content within mortar samples). The geometry of the rebars within the composites was all kept constant (precise to ±0.5 mm) to ensure that the positions of the rebars were not a contributing cause of variations (Fig. 1). 2.3. Characterisation A series of tests have been carried out to characterise the developed composites, and a summary of the test methods conducted and their purpose is identified in Table 2. 3. Results and discussions 3.1. Functional groups of sisal and its rebars for interface bonding
2.2. Sample preparation Dry sisal rope of 600 mm sizes was pre-stressed using a torsion machine of known pressure (20 rotational cycles). Known amounts of resin were thoroughly mixed and sonicated for approx. 10 min to remove most of the bubbles. The clear resin was then brushed uniformly onto the exposed surface of the sisal rope while it was under torsional stress for an application time of approx. 2 min (Fig. 1a). This resin coating keeps the fibres intact as well as keeping the overall rebar rigid and uniform, maintaining reliability. The resin is also used to protect the fibres from chemical attacks, prolonging their durability. Once the resin coating becomes sufficiently hard after about 24 h (at 22 C and 50% relative humidity), the ropes were cut to lengths of 150 mm in order to fit the prismatic moulds of 40 40 160 mm3. Samples were fitted with two tensile rebar reinforcements per mould using mesh spacers and silicone glue, such that a minimum of 5 mm cover is kept at the bottom and outer sides of the reinforcements (Fig. 1b). The sisal fibres were not treated in an effort to improve the interfacial strength between the
Fourier transform infrared spectroscopy (FTIR) was used to determine the functional groups of sisal fibres and their coated counterparts (rebars) (Fig. 2). A large quantity of OH functional groups was found within sisal fibres (Fig. 2a) as the 3290 cm1 peak denotes in Table 3. Sisal is a natural fibre, which is chemically comprised of cellulose, hemicellulose and lignin with impurities including pectin and wax. The presence of cellulose defines the chemical and physical properties of the fibre as it makes up the majority (70%) of the biological structure. Cellulose contains many chemical bonds that vary in type and extent compared to that of the synthetic resins. Hydroxyl is one example of these where its presence is particularly plentiful in sisal as the glucose chains are held together using hydrogen bonds between hydroxyl groups to form microfibrils. These OH bonds may be the primary reason for the strong interface bonding between the fibre and the matrix since strong hydrogen bonds are formed. Epoxy also has hydroxyl groups within the chemical structure which may also promote hydrogen bonding, however, the
Table 1 Typical properties of the various resins used.
Epoxy (E) Vinyl-ester (V) Polyester (P) Polyurethane (PU)
Resin:hardener by weight
Mix density (gcm3) @ 25 C [BS EN ISO 5350-Part B1]
Flexural strength (MPa) [BS EN ISO 178 [32]]
5:1 1:0.015 1:0.02 Part A 1:1.2 Part B
1.11 1.32 @20 C 1.09 1.03e1.08
81.42 153 78 e
46
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
Fig. 1. Prepared prismatic moulds with sisal-coated rebars fitted (a) and rebar geometry in the cementitious matrix (b).
Table 2 Destructive and non-destructive tests. Measurement Destructive Flexural property (Instron 5584, ±150 kN load cell) Moisture content and Water absorption
Non-destructive Scanning Electron Microscopy (SEM) Optical Microscopy (OM) Fourier Transform Infrared Spectroscopy (FTIR)
Method
Purpose
Samples of 40 40 160 mm3 were tested in threepoint bending at a loading rate of 3000 N/min. Rebar samples (∅3.5 50 mm) were oven dried at 60 C for 24 h before being sealed in distilled water baths for 98 h; samples were weighed after specified time intervals.
To observe flexural load-displacement behaviour, in compliance with BS EN 196-1:2005 [23]. To investigate the extent of water uptake by sisal fibres and to understand its influence on the fibre strength and durability.
Non-conductive samples of 40 15 15 mm were analysed at VP mode of 37 kPa. The same SEM samples were used to attain a visual colour image of the areas analysed. 50 scans were carried out per sample at a wavenumber range of 4000 to 1500 cm1.
To observe microstructural interactions between the material phases. To visually identify gaps/holes or other possible defects.
Fig. 2. FTIR spectra of the different resins and fibre.
To identify the functional groups and hence characterise particular compounds within the various materials.
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
47
Table 3 Band characteristics of FTIR spectra [33,34]. Characteristic Absorption (cm1) S
SE
SV
Region 1a 3290
3377
3331
SP
3060
3062
Assignment
Comment
3328 3430
O-H (m) N-H (m) ¼C-H and C¼C (m) C-H (w)
Intermol. hydrogen bonding, free water; cellulose, hemicellulose Urethane linkage [Ar-NHCOO] Within aromatic ring Asymmetric; epoxide
2970 2924 2857 2257
CeH (s-m) CeH (s-m) CeH (m) NCO (vs)
Asymmetric [ReCH3]; cellulose and hemicellulose Asymmetric [eCH2e(acyclic)] Symmetric [eCH2e(acyclic)] Asymmetric; isocyanate
C¼O (vs) C¼O (s) C¼C (s-m) ¼C-H and C¼C (v) ¼C-H and C¼C (v) In-plane N-H (s) ¼C-H and C¼C (v)
Aryl or a, b- unsat. (Ar-COOR); pectin, wax and hemicellulose, Alkyl urethane [R-NHCOO]; amide I Vinyl group Within aromatic ring; lignin [eCHeC¼CeCHe] aromatic Urethane linkage [Ar-NHCOO] Within aromatic ring; lignin
SPU
3038 Region 2 2921 2850
2963 2925 2873
2966 2929 2872
2963 2918 2852
Region 3 1729
1714
1714
1718 1692
1595
1607 1581
1503
1508
1637 1602
1599 1579 1527
1507
S ¼ Sisal, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S. a Band intensity: vs ¼ very strong, s ¼ strong, m ¼ medium, w ¼ weak, v ¼ variable.
prevalence of these hydroxyl group is considerably lower than that of the sisal (Fig. 2a). The intensity of the hydroxyl groups is also lower than the peaks for the other chemical groups within the epoxy structure, which is in contrast to that of sisal where the hydroxyl group has the highest absorption peak in its spectrum. All four resins are shown to have aromatic rings within their monomer units where polyester is found to have one such ring while the other three resins have two. The aromatic rings are responsible for enhancing the chemical/thermal stability of structures, thus, polyester is prone to corrosion/chemical shrinkage due to the exothermic and caustic fresh mortar in comparison to the other three resins accordingly [24]. This is evidenced in Fig. 4 where the polyester coating is shown to be mostly degenerated. The functional groups of the treated sisal rebars (orange curves) largely seem to resemble those of the coated resins (blue curves) but the intensity of the spectra peaks is reduced significantly (Fig. 2) except for polyester (Fig. 2d) where the two lines largely converge. This may be because much of the functional groups undergo a reaction and form bonds with the sisal fibre, thereby limiting their presence as the groups in the resin. The spectrum for epoxy coated sisal rebar (SE) has peaks that closely resemble those of the epoxy spectrum (Fig. 2a), except that the spectrum contains a peak at 1714 cm1 in the third region, corresponding to C¼O stretching vibration, which does not exist in epoxy but in the sisal spectrum. The polyurethane spectrum peak, 3328 cm1, in the region of the hydroxyl peak, however, this peak is primarily a result of the NH group (which can also form hydrogen bonds) within the polyurethane chemical structure. The polyester spectrum, like the polyurethane, does not have a hydroxyl peak. Good adhesion is primarily a result of the hydroxyl functional group and thus, one would expect this property of epoxy and vinylester which are found to contain this group as shown in the spectra. Polyurethane contains the group NH that also promotes adhesion in the compound, albeit to a reduced extent, as the nitrogen is also capable of forming hydrogen bonds similar to the oxygen in hydroxyl.
3.2. Effect of the behaviour of sisal rebar on interface Two issues that arise with the use of natural fibres are their ability to absorb moisture, which alters their strength and durability, and the moisture-induced swelling/shrinkage which is a
major cause of early cracking in concrete prior to loading. As discussed above, the peak at 3290 cm1, corresponding to eOH, gives rise to the ability of sisal to absorb moisture. At the initial period of the matrix setting, natural fibres could absorb the excess water from the cementitious mixture since the surrounding water concentration is greater than that of the moisture concentration within the fibres, and hence swell (enlarge the size of fibre diameter). The absorption of moisture to natural fibre means that the cementitious mixtures may not have sufficient water for complete hydration at the early stage of the hydration process or a higher water to cement ratio should be taken into account for the water absorbed by the natural fibre. At a later stage of curing, the moisture content within the fibres may dry out and thus, fibres shrink to a smaller size than that during the curing process of mortar, as shown in Fig. 3a, while the surrounded mortar may have initially set. The shrinkage of the fibre forms a gap between the fibre and mortar, which would infer that the bonding between both materials has diminished or weakened depending on the level of the shrinking effect. Further hydration of mortar and drying of composites may result in further shrinkage of sisal fibres, which could be higher than that of its surrounding mortar by considering that the rigid structure of mortar is formed. Such the gap between the fibre surface and mortar could reach to a gap size of Gw (Fig. 3a). The extent to which sisal could be affected by the absorption and shrinkage in the cementitious matrix was investigated via the analysis of microscopic images (Fig. 3b). It can be seen that the individual twine fibres have separated as well as the separation of the three twined sisal plies due to the drying and shrinkage of sisal fibre rebar. The colour change of the cementitious matrix surrounded rebar was also observed, which explains at later stage the movement of water from the fibres to the surrounding matrix, making the surrounding cementitious matrix shell damp. The damp mortar not only reduces its ability to protect the fibre but also compromises the strength of the composite. This is because the interfacial transition zone has deteriorated as a result of increased water content surrounding the fibre. Gaps can also be seen between the fibre and matrix, which would weaken the interaction and thus reduce the tension and compression relationship between the fibre and cementitious matrix. This has been reflected in the testing of its flexural strength (Fig. 3c). As seen from the graph, the load initially
48
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
Fig. 3. Varying diameters of the fibre as moisture content changes (a), a corresponding microscopic image of sisal reinforced cementitious matrix (b), stress-strain relationship of sisal reinforced cementitious matrix (c).
taken up by the cementitious matrix is not later supported by the sisal rebar to a significant extent after the cracking of the matrix. This investigation has helped observe the necessary improvements needed to further advance the fibre-matrix interaction and thus its strength [25]. To address this issue, four types of resin are applied to the sisal fibre bars. 3.3. Effect of various coated sisal fibres on interface Coating fibres with resin generates a significantly more compact structure (Fig. 4). The morphology from optical microscopy is given on the left column and those from SEM in the middle column with the interface between fibre and matrix being emphasised in the right column. Taking a comparison of Fig. 4S with Fig. 4SE as an example, it is apparent that coating natural fibres gives rise to a significant improvement in the bonding and fibre-matrix interface with possible reduction in fibre degradation [26]. This can be attributed to several factors: Firstly, as afore-discussed, coating natural fibre has resulted in a significant reduction of the intensity of eOH (Fig. 2a), which means a reduction in the water absorption. Furthermore, the resin coating also reduces open porosity (a network of open pores) between the fibre and cement-based matrix, thereby significantly decreasing the permeability with respect to water intake. This can be confirmed by the water absorption test (Table 4). One hour soaking of fibre in water resulted in the absorption of water 64% for uncoated sisal but only 1% for epoxy-coated sisal. The reduction in water absorption in the resin coated fibres causes less swelling of the natural fibre when mixed with the cementitious matrix. This causes a reduction in the formation of voids between the different components and thus produces a more compact interface. The reduction in water sorption also means less water movement from the natural fibre and the matrix and ensures a lower deterioration of the interfacial transition region as the cement is set. Secondly, the coating has altered the surface functionalities of natural fibres due partly to the prior reaction of natural fibre with the polymer, such as the reduction in C¼O 1729 cm1 when the resins were introduced (Fig. 2 and Table 3). The reduction or removal of low molecule functional groups would be able to improve the bonding of cement and rebars as those low molecule compounds usually retard the cement hydration [27] when they are not extracted and contact the cement matrix. Thirdly, an initial sorption capacity of rebars may also influence the bonding and interface integrity of sisal-cement matrix. It can be seen that different coatings have given rise to the different initial moisture within the rebars (Fig. 5). Sisal without coating apparently has a high capacity to store water within the fibre. Coating the fibre with a resin reduces this ability considerably as the fibre has less access to the atmosphere. Therefore, if the natural fibres are not
dried before coating, the embedded moisture within rebars could result in an adverse effect on the property of the composites. More moisture usually promotes higher shrinkage and movement of moisture from fibre to matrix. Scrutiny of Fig. 4 indicates that the different surface coating of natural fibres has a different effect on the bonding and interface structure. Fig. 4SE is the cross section of the epoxy coated sisal (SE) fibre composite and demonstrates that the sisal fibre is well integrated within the epoxy. It is apparent that the interface of cementSE has a low extent of major voids between the fibre and matrix, which is even lower than that of the steel composite (Fig. 4St). It should be noted that some of the composites show a small divergence between the sisal fibres and the resin as well as the cement and the resin, as seen within the SE composite. This divergence may be due to the epoxy resin and the sisal shrinking after moisture evaporation, which is highlighted by Fig. 5 where the resins and fibre experience significant moisture loss after curing. The maximum size of this divergence between the resin-matrix interface is about 10 mm while that of the resin-fibre interface can reach up to about 20 mm, as can be seen by the SEM images in Fig. 4. The compactness of the composite seems dependent on the type of resin coating used since Epoxy coating (SE) seems more compact than the Polyurethane coated sisal (SPU), which in turn seems more compact than the Vinylester (SV) and the Polyester coated (SP) natural fibre composites, respectively (Fig. 4). It is evident that the SP composite has an unusually large amount of voids between the fibre-matrix interface compared to SPU composites. The SP fibre may be more hydrophilic than other polymers used as the structure contains many oxygen atoms which can form hydrogen bonds with water molecules in the fresh mortar and, therefore, may absorb more water on the surface. This increase in absorption may cause these large voids when the moisture is evaporated which leads to much weaker mechanical properties. The SEM image of the SV composite is similar to that of the SP, albeit, with smaller voids. This suggests it is less hydrophilic compared to that of SP. It may also explain why it has better mechanical properties when compared to SP composite, which is discussed in the next sections. It is worth noting that during coating, the resin may not be able to fully seep into space between the three yarns due to the prestressing (twinning) of the yarns. Before the fibre shrunk due to moisture evaporation, the three yarns acted as a barrier to the resin coating. Drying the fibre before adding the resin coating prevents shrinkage, enabling maximum adhesion between the interfaces, and allowing the resin to seep within the three yarns more easily. Another possible improvement in composite mechanical properties is to apply resin to the individual yarns whilst pre-stressing (twinning) the yarns together (before resin polymerisation). Moreover, the addition of yarns helps add a textured surface, increasing the surface area at the interface, thereby increasing the
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
49
Fig. 4. OM & SEM images of composites: St ¼ Steel fibre, S ¼ Sisal natural fibre, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
50
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
Table 4 Average water absorption of natural fibre rebar (%). Time (h)
S
SE
SV
SP
SPU
1 2 4 24 28 98
64 65 65 65 65 65
1 1 1 3 4 11
7 6 7 11 11 14
9 9 9 11 11 13
1 2 2 5 6 15
S ¼ Sisal, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
Fig. 5. Moisture loss of preconditioned fibre rebar after oven drying (error bars represent standard deviation values of four replicas). S ¼ Sisal fibre, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
chemical attraction between the two materials as well as increasing physical interaction due to increased surface roughness. This makes it more resistant to separation. Furthermore, the morphology of the individual sisal fibres may be of significance in the fibre-matrix bond strength, as well as increasing the fibre content to introduce further deflection hardening before failure [28,29]. 3.4. Failure mechanism of coated natural fibre-inorganic matrix composites The coated natural fibre cementitious composites demonstrate complex behaviour of fluctuation in the characteristic stress-strain relationship (Fig. 6). Each of the composites exhibits two initial dips, which may indicate the formation of cracks within the samples. The first dips may reflect the failure of the cementitious matrix before the resin resists the load, which is the next layer of reinforcement (Fig. 6). The second cracks reflect the failure of the resin before the sisal fibre resists the load. It must be noted that the resin and mortar surrounding the crack regions may continue to sustain some of the load after cracking, as the remaining structure is capable of doing so [30]. This is represented by the few microcracks within the strain hardening region of the curves, generated under loading, before fracture was promoted by main crack [29,31]. It can be seen that the first dips in the graphs of the composites are roughly at the same strain because the cement matrices are all the same type, including the plain mortar sample (M-curve). The small variations may be attributed to different coatings which may also affect the matrix at the interface region. However, the depth of first dips vary with the type of coatings; those of the epoxy (E) and polyurethane (PU) coated composites are smaller than those of polyester (P) and vinylester (V) coated composites. This depth highlights an important property of the coatings that may have affected the interface and hence mechanical properties of the
composites. These characteristics are consistent with the morphology of composites (Fig. 4), which show less compact interface of SP-matrix and SV-matrix than those of SPU-matrix and SE-matrix composites. This is also confirmed by the fact that the stress at the first dip for mortar matrix is the highest as without the effect of coating-cement interface. This means that once the cementitious matrix fails during the testing, the load will quickly be conveyed further inside the composite, e.g. the interface of natural fibre-matrix. Less compact structure of interface may not be able to resist the load sufficiently. The second dips may be related to the capacity of the surface coating of natural fibre rebar and the characteristics of coatingnatural fibre interface (Fig. 6). It is very interesting that although the strain of SP- and SV-matrix is much higher than that of SPUand SE-matrix, the maximum stress for all four rebar matrices seems to be very similar. This may reflect the influence of the natural fibre-coating interface, since thin layers of resin coating may not significantly contribute to the mechanical properties. This is in addition to the role played by the inherent strength of the individual polymer. The denser morphology of SE and SPU compared to the SV and SP interface is one contributing factor to the discrepancy in strength. Another possible factor is elongation or microcracks of the coating layers. The rise in stress that follows the second dips demonstrates the combined effect of cement matrix, coating polymer and natural fibres (Fig. 6). In addition to the factors discussed above, the gaps formed within the architected yarns may play a role. During the coating process, some of the resins naturally seeped into the gaps within the fibre structure as shown in SE and SPU in comparison to SV and SP. This seepage could be affected by the type of polymers used. The load capacity of the coated rebars may be different due to the type of resin contained within the fibre, even though they are all the same type of fibre. It is worth to note that the SPU and the SE-matrix composites also experience an additional dip before a complete failure (Fig. 6). These dips may be due to further deterioration of polymer within the gaps of yarn structure, especially once they connect with the initial crack. 3.5. Mechanical properties of coated natural fibre rebar cementitious composites The composites vary in mechanical properties when analysing individual materials within the composite. The plain mortar sample was found to be stiffer than the cementitious matrix of the composites while all failed in a brittle manner. The composites seem to be coupled into two groups; with composites containing polyester and vinylester being tougher, while polyurethane and epoxy composites are stiffer (Table 5 and Fig. 6). The curves that corresponds to the mechanical properties of the resins display brittle behaviour, since they are all thermoset plastics which are used to hold the prestressed fibre in shape. The region of the graph corresponding to the epoxy coated fibre has a very small elastic region compared to the other composites with a large plastic region. Polyurethane fibre composite is the strongest and stiffest while the epoxy fibre composite is the second strongest (Fig. 6). All of the fibre reinforced composites display high toughness and provide a degree of flexibility and plasticity to the composites indicated by the large plastic regions. This is a beneficial mechanical property for structural materials since it increases resistive capacity before failing, allowing a structure to continue acting as a support under large forces [11]. This ductility could provide some sufficient warning of impending failure if the structure were to experience structural fatigue or an earthquake. In addition, the SE curve in the graph suggests the occurrence of necking at around 4 MPa before stabilising and this was not seen among the other composites (Fig. 6).
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
51
Fig. 6. Stress-strain relationship of cementitious composite flexure test at 7 days and the corresponding failure mechanism: Cm is the stress that the cementitious matrix can withstand before failure and Cr is the stress increase that the resin can resist before reaching point of failure. Cross section A-A depicts main crack perpendicular to the reinforcement at midpoint. M ¼ plain Mortar, SE ¼ Epoxy coated Sisal (S) reinforcing M, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
Table 5 Mean flexural strength of the various composites (MPa).a Days
M
S
SE
SV
SP
SPU
St
7 14 28
4.03 (5.28) 5.78 (6.96) 6.03 (8.36)
e e 6.09 (1.74)
12.39 (9.09) 18.66 (18.23) 16.47 (9.37)
10.88 (21.02) 17.9 (8.43) 18.39 (7.33)
9.5 (41.43) 14.48 (23.65) 14.51 (6.44)
14.35 (9.57) 19.68 (2.62) 19.8 (9.17)
e e 21.18 (4.47)
M ¼ Mortar, S ¼ Sisal, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S, St ¼ Steel fibre. a Value in ( ) is coefficient of variance (CV) in %.
It is apparent that the flexural strength varies between various combinations (Table 5). The variation of strength with curing times is because the cementitious matrix takes time to set. A scrutiny of the results shows that the flexural strength of SE- and PU-matrix composites is higher than that of SP- and SV-matrix composites, especially in the early stages of curing; however, this may not be accurate due to the high coefficient of variance. This is in agreement with different quality of the interface, as discussed in section 3.2, which is probably due to the compatibility and level of reaction between the polymer and natural fibres. This is also the case for the functional groups existed on the natural fibre and polymer surface, as discussed in section 3.1. All natural rebar composites are stronger than cement-based composites reinforced with sisal fibres (S), which is found to be the weakest (Table 5 and Fig. 3c). This highlights the significant benefit of using resin coating for natural fibre rebar as a dramatic increase in performance is witnessed. As seen, the maximum strength of SPU or SV reinforced composites is more than 3 times that of sisal fibre reinforced composites (Table 5 and Fig. 3c). Although the steel reinforced composite is the strongest, the strength at 28 days for the steel composites is only about 7% higher than that of SPU composites. This suggests that the widespread use of steel reinforcement in contemporary society may be reduced in some structural applications by potential reinforcements of comparable strength such as polyurethane (PU) coated natural fibre.
composites have been thoroughly investigated to determine the bonding and failure mechanisms. Some specific conclusions can be drawn as follows:
4. Conclusions
References
Natural fibre rebars have been developed and their use as reinforcement in structural concrete members has been tested in this paper. The behaviour and interface of the developed
1) A less compact interface was found when bare natural fibre rebar composite was used in comparison to coated natural fibre rebar composites, resulting in lower mechanical properties; 2) When the natural fibre is coated with certain resins such as polyurethane, the composite could attain mechanical (e.g. flexure) strength rivalling that of conventional steel reinforced composites; 3) All coatings on sisal rebars resulted in significant improvement in bonding interface and mechanical properties of the composites, albeit to varying degrees. It appears that the structure and quality of the resin coating influenced the strength of the composites depending on how well it protected the fibres from moisture; 4) The structural members with coated natural fibre rebars went through three stages of failure process with two clear dips in the stress-strain plots: the first dip occurred at similar flexure strains of approximately 10% regardless of the composite type, while the second dip originates from the mechanical failure of the resin coating with flexure strains reaching up to about 30% depending on the type of resin coating applied. Load capacity after the second dip depended on the architecture of the rebars.
[1] R.S. Blackburn, Biodegradable and Sustainable Fibres, Woodhead Publishing Limited, Cambridge, 2005. [2] A. Bentur, S. Mindess, Fibre Reinforced Cementitious Composites, second ed., Taylor & Francis, London and New York, 2007.
52
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
[3] M. Fan, Characterization and performance of elementary hemp fibres: factors influencing tensile strength, BioResources 5 (2010) 2307e2322. [4] T. Gurunathan, S. Mohanty, S.K. Nayak, A review of the recent developments in biocomposites based on natural fibres and their application perspectives, Compos. Part A Appl. Sci. Manuf. 77 (2015) 1e25. [5] A. Mohanty, M. Misra, L. Drzal, Natural Fibers, Biopolymers, and Biocomposites, CRC Press, Boca Raton, 2005. €m, L.A. Berglund, R.D.T. Filho, Morphology and me[6] K. Oksman, L. Wallstro chanical properties of unidirectional sisal- epoxy composites, J. Appl. Polym. Sci. 84 (2002) 2358e2365. [7] P. Wambua, J. Ivens, I. Verpoest, Natural fibres: can they replace glass in fibre reinforced plastics? Compos. Sci. Technol. 63 (2003) 1259e1264. [8] X. Zhou, S.H. Ghaffar, W. Dong, O. Oladiran, M. Fan, Fracture and impact properties of short discrete jute fibre-reinforced cementitious composites, Mater. Des. 49 (2013) 35e47. [9] Food and Agricultural Commodities Production/Countries by Commodity, FAOSTAT, 2013 (fao.org). [10] Y. Li, Y. Mai, L. Ye, Sisal fibre and its composites: a review of recent developments, Compos. Sci. Technol. 60 (2000) 2037e2055. [11] A. Belaadi, A. Bezazi, M. Bourchak, F. Scarpa, Tensile static and fatigue behaviour of sisal fibres, Mater. Des. 46 (2013) 76e83. [12] C. Stevens, J. Müssig, Industrial Applications of Natural Fibres: Structure, Properties and Technical Applications, John Wiley & Sons, Bremen, 2010. [13] O. Onuaguluchi, N. Banthia, Plant-based natural fibre reinforced cement composites: a review, Cem. Concr. Compos 68 (2016) 96e108. [14] F. Gentles, Characterisation of Environmentally Friendly Fibres for Composites, University of Strathclyde, 2015. [15] F. Matthews, R. Rawlings, Composite Materials: Engineering and Science, Woodhead Publishing Limited, London, 1999. [16] BS EN 197-1:2011, Cement Part 1: Composition, Specifications and Conformity Criteria for Common Cements, BSI Stand., 2011. [17] J. Gassan, A. Bledzki, Possibilities for improving the mechanical properties of jute/epoxy composites by alkali treatment of fibres, Compos. Sci. Technol. 59 (1999) 1303e1309. [18] M. Rong, M. Zhang, Y. Liu, G. Yang, H. Zeng, The effect of fiber treatment on the mechanical properties of unidirectional sisal-reinforced epoxy composites, Compos. Sci. Technol. 61 (2001) 1437e1447. [19] T. Doan, H. Brodowsky, E. M€ ader, Jute fibre/epoxy composites: surface
[20] [21]
[22] [23] [24] [25]
[26]
[27] [28]
[29]
[30] [31]
[32] [33] [34]
properties and interfacial adhesion, Compos. Sci. Technol. 72 (2012) 1160e1166. W. Liu, T. Xie, R. Qiu, M. Fan, N-methylol acrylamide grafting bamboo fibers and their composites, Compos. Sci. Technol. 117 (2015) 100e106. K.L. Pickering, M.G. Aruan Efendy, T.M. Le, A review of recent developments in natural fibre composites and their mechanical performance, Compos. Part A Appl. Sci. Manuf. 83 (2015) 98e112. M. Ali, X. Li, N. Chouw, Experimental investigations on bond strength between coconut fibre and concrete, Mater. Des. 44 (2013) 596e605. BS EN 196-1:2005, Methods of Testing Cement d Part 1: Determination of Strength, BSI Stand, 2005. S. Sen, S. Patil, D.S. Argyropoulos, Thermal properties of lignin in copolymers, blends, and composites: a review, Green Chem. 17 (2015). K.L. Pickering, M.G.A. Efendy, T.M. Le, A review of recent developments in natural fibre composites and their mechanical performance, Compos. Part A Appl. Sci. Manuf. 83 (2016) 98e112. M.E.A. Fidelis, R.D. Toledo Filho, F. de A. Silva, V. Mechtcherine, M. Butler, S. Hempel, The effect of accelerated aging on the interface of jute textile reinforced concrete, Cem. Concr. Compos 74 (2016) 7e15. J. Hwang, Effects of Cement Treatment on the One-dimensional Consolidation Behavior of a Highly Organic Soil, Purdue University, 2006. F.D.A. Silva, B. Mobasher, C. Soranakom, R.D.T. Filho, Effect of fiber shape and morphology on interfacial bond and cracking behaviors of sisal fiber cement based composites, Cem. Concr. Compos 33 (2011) 814e823. a, P.R.L. Lima, Finite strip K. De Jesus Nagahama, A. De Souza Matos Gade modeling of cementitious laminates reinforced with sisal fibers, Cem. Concr. Compos 63 (2015) 8e16. F. de A. Silva, B. Mobasher, R.D.T. Filho, Cracking mechanisms in durable sisal fiber reinforced cement composites, Cem. Concr. Compos 31 (2009) 721e730. J.D.A.M. Filho, F.D.A. Silva, R.D. Toledo Filho, Degradation kinetics and aging mechanisms on sisal fiber cement composite systems, Cem. Concr. Compos 40 (2013) 30e39. BS EN ISO 178:2010þA1:2013, Plastics d Determination of Flexural Properties, 2011. G. Socrates, Infrared and Raman Characteristic Group Frequencies: Tables and Charts, John Wiley & Sons, London, 2004. M. Hesse, H. Meier, B. Zeeh, Spektroskopische Methoden in der organischen Chemie, eighth ed., Thieme, 2014.
Contents lists available at ScienceDirect
Cement and Concrete Composites journal homepage: www.elsevier.com/locate/cemconcomp
Interfacial properties and structural performance of resin-coated natural fibre rebars within cementitious matrices Hassan Ahmad, Mizi Fan* Nanocellulose and Biocomposites Research Centre, College of Engineering, Design and Physical Sciences, Brunel University London, UB8 3PH, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 October 2016 Received in revised form 24 November 2017 Accepted 4 December 2017 Available online 6 December 2017
This paper investigates the interface and behaviour of continuous natural fibre rebar reinforced cementitious composites. Various resins were used to coat sisal fibres which were then used as reinforcement within the composites. Non-destructive tests (SEM, OM, FTIR) were employed to evaluate the interfacial bonding and interaction between the sisal, resin and cementitious matrix. The results showed that the uncoated fibre could not develop a compact interface due to the moisture absorption/desorption of the fibre, which caused low mechanical properties of the composite. The coating was able to improve the strength of the rebars and reduce the OH concentration, which improved the interfacial bonding and integrity. The flexural stress-strain relation of the composites exhibited three regions with two clear dips, reflecting the progressive stress transfer and failure of the composite constituents and their interactions. The comparable mechanical properties to those of steel rebar reinforcement demonstrate the potential of the resin-coated sisal fibre rebar for structural applications. © 2017 Published by Elsevier Ltd.
Keywords: Sisal fibre Resin coating/surface treatment Natural fibre rebar cementitious composite Fibre/matrix bond Interface Microstructural analysis
1. Introduction Various forms of fibres have been developed to overcome the limitations of concrete as a quasi-brittle material that is weak in tension. Fibres typically act to restrict crack growth within concrete. Natural fibres are traditionally classified into four categories according to their origin: stem/bast, leaf, fruit/surface and wood/ cellulose [1]. Leaf fibres provide the fundamental rigidity and strength of leaves and are extracted by means of scraping and drying crushed leaves. In comparison to stem/bast fibres, the leaf fibres are commonly tougher, rougher in texture and stronger on average. For example, the strength and stiffness of sisal fibre range from 550 to 750 MPa and 17e38 GPa compared to those of jute fibre from 300 to 800 MPa and 10e30 GPa respectively [2e8]. Sisal fibres, being the most common type of leaf fibres, are produced from various varieties of the agave plant, which is indigenous to the arid regions of Africa and South America, mainly Brazil e the biggest commercial producer of Agave sisalana amounting to an estimate of 150,000 tonnes [9]. Known to have one of the strongest natural fibres, the agave plant is characterised by its
* Corresponding author. E-mail addresses: [email protected] (H. Ahmad), [email protected]. uk (M. Fan). https://doi.org/10.1016/j.cemconcomp.2017.12.002 0958-9465/© 2017 Published by Elsevier Ltd.
leaves, which grow to a length of over one metre, and yield a long creamy white and very strong fibre. Sisal is a fast-growing plant that remains reproductive for 10e12 years and produces 180 to 240 leaves in a lifetime. In addition, having a total annual production of approximately 300,000 tonnes with a low-cost production, sisal fibre is increasingly being used in the building industry, particularly for plaster reinforcement. As sisal has been used for many years for reinforcing gypsum products, there is an emerging potential for sisal fibres being used to reinforce other composites for low-cost housing applications, namely reinforcing concrete matrices [10e13]. The characteristics and performance of the fibre-reinforced composites depend on the properties of the individual components within the composites including the fibre, the individual matrices and the interfaces between them. One must, therefore, investigate the nature of each of these components to gain a comprehensive understanding of the properties of the final composite. When considering vegetable fibre reinforced composites, the major research challenge has been trying to achieve a strong interface, especially a dense interfacial transition zone, within the composites. However, the surface chemistry of natural fibres and the negative effects of moisture diffusion between the fibre and the inorganic-bonded-matrix (e.g. mortar) on the interfacial bonding complicate this. In realising and optimising the requirements for
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
exceptional bonding mechanism, i.e. mechanical, electrostatic, interdiffusion or chemical bonding between the fibre and matrix [14], the fibre and matrix can such reach their individual strength such that the load applied is uniformly distributed within the lattice structure [15]. Moreover, for natural fibre-inorganic composites, there exists an additional major challenge: hydrophilic nature of natural fibre complicating with the essential presence of moisture for the inorganic matrix (requirement for curing). Hence, this paper is to systematically analyse and characterise the interfacial systems of the natural fibre (sisal fibre as an example) e inorganic matrix (cement-based as a candidate) composites (FRC). The findings of this research will result in a better understanding of the interface and interaction of natural fibre with the cementitious constituent, provide fundamental information for further development of the natural fibre inorganic composites, and such facilitate and promote the use of natural fibres as a structural reinforcement for cementitious composites that could be a viable alternative to steel in some applications. 2. Experimental plan 2.1. Raw materials Natural fibre sisals (Agave sisalana) were purchased online in a form of ropes comprised of 3-ply twined yarns to make up 3 mm diameter string of the desired length. It is graded as type-1 sisal, which is the finest known quality. Additionally, four types of resins (slow cure time) were bought from a commercial supplier for coating sisal fibres; epoxy, polyurethane, vinylester and polyester. The properties of the resins can be summarised in Table 1. 3 mm marine grade steel wire rope was used as a comparison. Standard CEM II Portland-limestone (~6e20%) cement (Class 32,5 R), complying with BS EN 197-1 [16] was used for making the matrix along with sharp sand bought from a commercial building material supplier.
45
fibre and resin, the effectiveness of which has been proven by some studies [17e22]. Samples of uncoated steel rebar were fitted in the same way. The mix proportion used for making the cementitious matrix used was a 1:1.5 ratio of cement to sand with water to cement (w/c) ratio of 0.4. The sand used was oven dried at 100 C for 24 h and sieved to a particle size distribution of 1 mm. The fresh mortar was fabricated using a bench-top mixer before casting into the moulds and compacting for 1 min using a vibrating table, conforming to EN 196-1 [23]. The moulds were then covered with a plastic film and water sprayed on top of the plastic film to protect and control the moisture loss. After 24 h, the cementitious composites were de-moulded and left to cure at room temperature and pressure of 22 ± 1 C and ~100 kN/m2 with relative humidity of 50% within a curing chamber till the ages of 7, 14 and 28 days. To improve the reliability of the investigation, four samples of each type of composite were tested and the mean values were used. Control variables were maintained, such as keeping the proportions of the ingredients the same and applying the same amount of resin used on the rebar (6 g of any resin applied for every 60 cm length of twined sisal rope), such that coated sisal rebars have diameters of 3.5 ± 0.2 mm when measured via a Vernier caliper (~0.3 ± 0.02 vol% of fibre content within mortar samples). The geometry of the rebars within the composites was all kept constant (precise to ±0.5 mm) to ensure that the positions of the rebars were not a contributing cause of variations (Fig. 1). 2.3. Characterisation A series of tests have been carried out to characterise the developed composites, and a summary of the test methods conducted and their purpose is identified in Table 2. 3. Results and discussions 3.1. Functional groups of sisal and its rebars for interface bonding
2.2. Sample preparation Dry sisal rope of 600 mm sizes was pre-stressed using a torsion machine of known pressure (20 rotational cycles). Known amounts of resin were thoroughly mixed and sonicated for approx. 10 min to remove most of the bubbles. The clear resin was then brushed uniformly onto the exposed surface of the sisal rope while it was under torsional stress for an application time of approx. 2 min (Fig. 1a). This resin coating keeps the fibres intact as well as keeping the overall rebar rigid and uniform, maintaining reliability. The resin is also used to protect the fibres from chemical attacks, prolonging their durability. Once the resin coating becomes sufficiently hard after about 24 h (at 22 C and 50% relative humidity), the ropes were cut to lengths of 150 mm in order to fit the prismatic moulds of 40 40 160 mm3. Samples were fitted with two tensile rebar reinforcements per mould using mesh spacers and silicone glue, such that a minimum of 5 mm cover is kept at the bottom and outer sides of the reinforcements (Fig. 1b). The sisal fibres were not treated in an effort to improve the interfacial strength between the
Fourier transform infrared spectroscopy (FTIR) was used to determine the functional groups of sisal fibres and their coated counterparts (rebars) (Fig. 2). A large quantity of OH functional groups was found within sisal fibres (Fig. 2a) as the 3290 cm1 peak denotes in Table 3. Sisal is a natural fibre, which is chemically comprised of cellulose, hemicellulose and lignin with impurities including pectin and wax. The presence of cellulose defines the chemical and physical properties of the fibre as it makes up the majority (70%) of the biological structure. Cellulose contains many chemical bonds that vary in type and extent compared to that of the synthetic resins. Hydroxyl is one example of these where its presence is particularly plentiful in sisal as the glucose chains are held together using hydrogen bonds between hydroxyl groups to form microfibrils. These OH bonds may be the primary reason for the strong interface bonding between the fibre and the matrix since strong hydrogen bonds are formed. Epoxy also has hydroxyl groups within the chemical structure which may also promote hydrogen bonding, however, the
Table 1 Typical properties of the various resins used.
Epoxy (E) Vinyl-ester (V) Polyester (P) Polyurethane (PU)
Resin:hardener by weight
Mix density (gcm3) @ 25 C [BS EN ISO 5350-Part B1]
Flexural strength (MPa) [BS EN ISO 178 [32]]
5:1 1:0.015 1:0.02 Part A 1:1.2 Part B
1.11 1.32 @20 C 1.09 1.03e1.08
81.42 153 78 e
46
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
Fig. 1. Prepared prismatic moulds with sisal-coated rebars fitted (a) and rebar geometry in the cementitious matrix (b).
Table 2 Destructive and non-destructive tests. Measurement Destructive Flexural property (Instron 5584, ±150 kN load cell) Moisture content and Water absorption
Non-destructive Scanning Electron Microscopy (SEM) Optical Microscopy (OM) Fourier Transform Infrared Spectroscopy (FTIR)
Method
Purpose
Samples of 40 40 160 mm3 were tested in threepoint bending at a loading rate of 3000 N/min. Rebar samples (∅3.5 50 mm) were oven dried at 60 C for 24 h before being sealed in distilled water baths for 98 h; samples were weighed after specified time intervals.
To observe flexural load-displacement behaviour, in compliance with BS EN 196-1:2005 [23]. To investigate the extent of water uptake by sisal fibres and to understand its influence on the fibre strength and durability.
Non-conductive samples of 40 15 15 mm were analysed at VP mode of 37 kPa. The same SEM samples were used to attain a visual colour image of the areas analysed. 50 scans were carried out per sample at a wavenumber range of 4000 to 1500 cm1.
To observe microstructural interactions between the material phases. To visually identify gaps/holes or other possible defects.
Fig. 2. FTIR spectra of the different resins and fibre.
To identify the functional groups and hence characterise particular compounds within the various materials.
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
47
Table 3 Band characteristics of FTIR spectra [33,34]. Characteristic Absorption (cm1) S
SE
SV
Region 1a 3290
3377
3331
SP
3060
3062
Assignment
Comment
3328 3430
O-H (m) N-H (m) ¼C-H and C¼C (m) C-H (w)
Intermol. hydrogen bonding, free water; cellulose, hemicellulose Urethane linkage [Ar-NHCOO] Within aromatic ring Asymmetric; epoxide
2970 2924 2857 2257
CeH (s-m) CeH (s-m) CeH (m) NCO (vs)
Asymmetric [ReCH3]; cellulose and hemicellulose Asymmetric [eCH2e(acyclic)] Symmetric [eCH2e(acyclic)] Asymmetric; isocyanate
C¼O (vs) C¼O (s) C¼C (s-m) ¼C-H and C¼C (v) ¼C-H and C¼C (v) In-plane N-H (s) ¼C-H and C¼C (v)
Aryl or a, b- unsat. (Ar-COOR); pectin, wax and hemicellulose, Alkyl urethane [R-NHCOO]; amide I Vinyl group Within aromatic ring; lignin [eCHeC¼CeCHe] aromatic Urethane linkage [Ar-NHCOO] Within aromatic ring; lignin
SPU
3038 Region 2 2921 2850
2963 2925 2873
2966 2929 2872
2963 2918 2852
Region 3 1729
1714
1714
1718 1692
1595
1607 1581
1503
1508
1637 1602
1599 1579 1527
1507
S ¼ Sisal, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S. a Band intensity: vs ¼ very strong, s ¼ strong, m ¼ medium, w ¼ weak, v ¼ variable.
prevalence of these hydroxyl group is considerably lower than that of the sisal (Fig. 2a). The intensity of the hydroxyl groups is also lower than the peaks for the other chemical groups within the epoxy structure, which is in contrast to that of sisal where the hydroxyl group has the highest absorption peak in its spectrum. All four resins are shown to have aromatic rings within their monomer units where polyester is found to have one such ring while the other three resins have two. The aromatic rings are responsible for enhancing the chemical/thermal stability of structures, thus, polyester is prone to corrosion/chemical shrinkage due to the exothermic and caustic fresh mortar in comparison to the other three resins accordingly [24]. This is evidenced in Fig. 4 where the polyester coating is shown to be mostly degenerated. The functional groups of the treated sisal rebars (orange curves) largely seem to resemble those of the coated resins (blue curves) but the intensity of the spectra peaks is reduced significantly (Fig. 2) except for polyester (Fig. 2d) where the two lines largely converge. This may be because much of the functional groups undergo a reaction and form bonds with the sisal fibre, thereby limiting their presence as the groups in the resin. The spectrum for epoxy coated sisal rebar (SE) has peaks that closely resemble those of the epoxy spectrum (Fig. 2a), except that the spectrum contains a peak at 1714 cm1 in the third region, corresponding to C¼O stretching vibration, which does not exist in epoxy but in the sisal spectrum. The polyurethane spectrum peak, 3328 cm1, in the region of the hydroxyl peak, however, this peak is primarily a result of the NH group (which can also form hydrogen bonds) within the polyurethane chemical structure. The polyester spectrum, like the polyurethane, does not have a hydroxyl peak. Good adhesion is primarily a result of the hydroxyl functional group and thus, one would expect this property of epoxy and vinylester which are found to contain this group as shown in the spectra. Polyurethane contains the group NH that also promotes adhesion in the compound, albeit to a reduced extent, as the nitrogen is also capable of forming hydrogen bonds similar to the oxygen in hydroxyl.
3.2. Effect of the behaviour of sisal rebar on interface Two issues that arise with the use of natural fibres are their ability to absorb moisture, which alters their strength and durability, and the moisture-induced swelling/shrinkage which is a
major cause of early cracking in concrete prior to loading. As discussed above, the peak at 3290 cm1, corresponding to eOH, gives rise to the ability of sisal to absorb moisture. At the initial period of the matrix setting, natural fibres could absorb the excess water from the cementitious mixture since the surrounding water concentration is greater than that of the moisture concentration within the fibres, and hence swell (enlarge the size of fibre diameter). The absorption of moisture to natural fibre means that the cementitious mixtures may not have sufficient water for complete hydration at the early stage of the hydration process or a higher water to cement ratio should be taken into account for the water absorbed by the natural fibre. At a later stage of curing, the moisture content within the fibres may dry out and thus, fibres shrink to a smaller size than that during the curing process of mortar, as shown in Fig. 3a, while the surrounded mortar may have initially set. The shrinkage of the fibre forms a gap between the fibre and mortar, which would infer that the bonding between both materials has diminished or weakened depending on the level of the shrinking effect. Further hydration of mortar and drying of composites may result in further shrinkage of sisal fibres, which could be higher than that of its surrounding mortar by considering that the rigid structure of mortar is formed. Such the gap between the fibre surface and mortar could reach to a gap size of Gw (Fig. 3a). The extent to which sisal could be affected by the absorption and shrinkage in the cementitious matrix was investigated via the analysis of microscopic images (Fig. 3b). It can be seen that the individual twine fibres have separated as well as the separation of the three twined sisal plies due to the drying and shrinkage of sisal fibre rebar. The colour change of the cementitious matrix surrounded rebar was also observed, which explains at later stage the movement of water from the fibres to the surrounding matrix, making the surrounding cementitious matrix shell damp. The damp mortar not only reduces its ability to protect the fibre but also compromises the strength of the composite. This is because the interfacial transition zone has deteriorated as a result of increased water content surrounding the fibre. Gaps can also be seen between the fibre and matrix, which would weaken the interaction and thus reduce the tension and compression relationship between the fibre and cementitious matrix. This has been reflected in the testing of its flexural strength (Fig. 3c). As seen from the graph, the load initially
48
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
Fig. 3. Varying diameters of the fibre as moisture content changes (a), a corresponding microscopic image of sisal reinforced cementitious matrix (b), stress-strain relationship of sisal reinforced cementitious matrix (c).
taken up by the cementitious matrix is not later supported by the sisal rebar to a significant extent after the cracking of the matrix. This investigation has helped observe the necessary improvements needed to further advance the fibre-matrix interaction and thus its strength [25]. To address this issue, four types of resin are applied to the sisal fibre bars. 3.3. Effect of various coated sisal fibres on interface Coating fibres with resin generates a significantly more compact structure (Fig. 4). The morphology from optical microscopy is given on the left column and those from SEM in the middle column with the interface between fibre and matrix being emphasised in the right column. Taking a comparison of Fig. 4S with Fig. 4SE as an example, it is apparent that coating natural fibres gives rise to a significant improvement in the bonding and fibre-matrix interface with possible reduction in fibre degradation [26]. This can be attributed to several factors: Firstly, as afore-discussed, coating natural fibre has resulted in a significant reduction of the intensity of eOH (Fig. 2a), which means a reduction in the water absorption. Furthermore, the resin coating also reduces open porosity (a network of open pores) between the fibre and cement-based matrix, thereby significantly decreasing the permeability with respect to water intake. This can be confirmed by the water absorption test (Table 4). One hour soaking of fibre in water resulted in the absorption of water 64% for uncoated sisal but only 1% for epoxy-coated sisal. The reduction in water absorption in the resin coated fibres causes less swelling of the natural fibre when mixed with the cementitious matrix. This causes a reduction in the formation of voids between the different components and thus produces a more compact interface. The reduction in water sorption also means less water movement from the natural fibre and the matrix and ensures a lower deterioration of the interfacial transition region as the cement is set. Secondly, the coating has altered the surface functionalities of natural fibres due partly to the prior reaction of natural fibre with the polymer, such as the reduction in C¼O 1729 cm1 when the resins were introduced (Fig. 2 and Table 3). The reduction or removal of low molecule functional groups would be able to improve the bonding of cement and rebars as those low molecule compounds usually retard the cement hydration [27] when they are not extracted and contact the cement matrix. Thirdly, an initial sorption capacity of rebars may also influence the bonding and interface integrity of sisal-cement matrix. It can be seen that different coatings have given rise to the different initial moisture within the rebars (Fig. 5). Sisal without coating apparently has a high capacity to store water within the fibre. Coating the fibre with a resin reduces this ability considerably as the fibre has less access to the atmosphere. Therefore, if the natural fibres are not
dried before coating, the embedded moisture within rebars could result in an adverse effect on the property of the composites. More moisture usually promotes higher shrinkage and movement of moisture from fibre to matrix. Scrutiny of Fig. 4 indicates that the different surface coating of natural fibres has a different effect on the bonding and interface structure. Fig. 4SE is the cross section of the epoxy coated sisal (SE) fibre composite and demonstrates that the sisal fibre is well integrated within the epoxy. It is apparent that the interface of cementSE has a low extent of major voids between the fibre and matrix, which is even lower than that of the steel composite (Fig. 4St). It should be noted that some of the composites show a small divergence between the sisal fibres and the resin as well as the cement and the resin, as seen within the SE composite. This divergence may be due to the epoxy resin and the sisal shrinking after moisture evaporation, which is highlighted by Fig. 5 where the resins and fibre experience significant moisture loss after curing. The maximum size of this divergence between the resin-matrix interface is about 10 mm while that of the resin-fibre interface can reach up to about 20 mm, as can be seen by the SEM images in Fig. 4. The compactness of the composite seems dependent on the type of resin coating used since Epoxy coating (SE) seems more compact than the Polyurethane coated sisal (SPU), which in turn seems more compact than the Vinylester (SV) and the Polyester coated (SP) natural fibre composites, respectively (Fig. 4). It is evident that the SP composite has an unusually large amount of voids between the fibre-matrix interface compared to SPU composites. The SP fibre may be more hydrophilic than other polymers used as the structure contains many oxygen atoms which can form hydrogen bonds with water molecules in the fresh mortar and, therefore, may absorb more water on the surface. This increase in absorption may cause these large voids when the moisture is evaporated which leads to much weaker mechanical properties. The SEM image of the SV composite is similar to that of the SP, albeit, with smaller voids. This suggests it is less hydrophilic compared to that of SP. It may also explain why it has better mechanical properties when compared to SP composite, which is discussed in the next sections. It is worth noting that during coating, the resin may not be able to fully seep into space between the three yarns due to the prestressing (twinning) of the yarns. Before the fibre shrunk due to moisture evaporation, the three yarns acted as a barrier to the resin coating. Drying the fibre before adding the resin coating prevents shrinkage, enabling maximum adhesion between the interfaces, and allowing the resin to seep within the three yarns more easily. Another possible improvement in composite mechanical properties is to apply resin to the individual yarns whilst pre-stressing (twinning) the yarns together (before resin polymerisation). Moreover, the addition of yarns helps add a textured surface, increasing the surface area at the interface, thereby increasing the
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
49
Fig. 4. OM & SEM images of composites: St ¼ Steel fibre, S ¼ Sisal natural fibre, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
50
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
Table 4 Average water absorption of natural fibre rebar (%). Time (h)
S
SE
SV
SP
SPU
1 2 4 24 28 98
64 65 65 65 65 65
1 1 1 3 4 11
7 6 7 11 11 14
9 9 9 11 11 13
1 2 2 5 6 15
S ¼ Sisal, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
Fig. 5. Moisture loss of preconditioned fibre rebar after oven drying (error bars represent standard deviation values of four replicas). S ¼ Sisal fibre, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
chemical attraction between the two materials as well as increasing physical interaction due to increased surface roughness. This makes it more resistant to separation. Furthermore, the morphology of the individual sisal fibres may be of significance in the fibre-matrix bond strength, as well as increasing the fibre content to introduce further deflection hardening before failure [28,29]. 3.4. Failure mechanism of coated natural fibre-inorganic matrix composites The coated natural fibre cementitious composites demonstrate complex behaviour of fluctuation in the characteristic stress-strain relationship (Fig. 6). Each of the composites exhibits two initial dips, which may indicate the formation of cracks within the samples. The first dips may reflect the failure of the cementitious matrix before the resin resists the load, which is the next layer of reinforcement (Fig. 6). The second cracks reflect the failure of the resin before the sisal fibre resists the load. It must be noted that the resin and mortar surrounding the crack regions may continue to sustain some of the load after cracking, as the remaining structure is capable of doing so [30]. This is represented by the few microcracks within the strain hardening region of the curves, generated under loading, before fracture was promoted by main crack [29,31]. It can be seen that the first dips in the graphs of the composites are roughly at the same strain because the cement matrices are all the same type, including the plain mortar sample (M-curve). The small variations may be attributed to different coatings which may also affect the matrix at the interface region. However, the depth of first dips vary with the type of coatings; those of the epoxy (E) and polyurethane (PU) coated composites are smaller than those of polyester (P) and vinylester (V) coated composites. This depth highlights an important property of the coatings that may have affected the interface and hence mechanical properties of the
composites. These characteristics are consistent with the morphology of composites (Fig. 4), which show less compact interface of SP-matrix and SV-matrix than those of SPU-matrix and SE-matrix composites. This is also confirmed by the fact that the stress at the first dip for mortar matrix is the highest as without the effect of coating-cement interface. This means that once the cementitious matrix fails during the testing, the load will quickly be conveyed further inside the composite, e.g. the interface of natural fibre-matrix. Less compact structure of interface may not be able to resist the load sufficiently. The second dips may be related to the capacity of the surface coating of natural fibre rebar and the characteristics of coatingnatural fibre interface (Fig. 6). It is very interesting that although the strain of SP- and SV-matrix is much higher than that of SPUand SE-matrix, the maximum stress for all four rebar matrices seems to be very similar. This may reflect the influence of the natural fibre-coating interface, since thin layers of resin coating may not significantly contribute to the mechanical properties. This is in addition to the role played by the inherent strength of the individual polymer. The denser morphology of SE and SPU compared to the SV and SP interface is one contributing factor to the discrepancy in strength. Another possible factor is elongation or microcracks of the coating layers. The rise in stress that follows the second dips demonstrates the combined effect of cement matrix, coating polymer and natural fibres (Fig. 6). In addition to the factors discussed above, the gaps formed within the architected yarns may play a role. During the coating process, some of the resins naturally seeped into the gaps within the fibre structure as shown in SE and SPU in comparison to SV and SP. This seepage could be affected by the type of polymers used. The load capacity of the coated rebars may be different due to the type of resin contained within the fibre, even though they are all the same type of fibre. It is worth to note that the SPU and the SE-matrix composites also experience an additional dip before a complete failure (Fig. 6). These dips may be due to further deterioration of polymer within the gaps of yarn structure, especially once they connect with the initial crack. 3.5. Mechanical properties of coated natural fibre rebar cementitious composites The composites vary in mechanical properties when analysing individual materials within the composite. The plain mortar sample was found to be stiffer than the cementitious matrix of the composites while all failed in a brittle manner. The composites seem to be coupled into two groups; with composites containing polyester and vinylester being tougher, while polyurethane and epoxy composites are stiffer (Table 5 and Fig. 6). The curves that corresponds to the mechanical properties of the resins display brittle behaviour, since they are all thermoset plastics which are used to hold the prestressed fibre in shape. The region of the graph corresponding to the epoxy coated fibre has a very small elastic region compared to the other composites with a large plastic region. Polyurethane fibre composite is the strongest and stiffest while the epoxy fibre composite is the second strongest (Fig. 6). All of the fibre reinforced composites display high toughness and provide a degree of flexibility and plasticity to the composites indicated by the large plastic regions. This is a beneficial mechanical property for structural materials since it increases resistive capacity before failing, allowing a structure to continue acting as a support under large forces [11]. This ductility could provide some sufficient warning of impending failure if the structure were to experience structural fatigue or an earthquake. In addition, the SE curve in the graph suggests the occurrence of necking at around 4 MPa before stabilising and this was not seen among the other composites (Fig. 6).
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
51
Fig. 6. Stress-strain relationship of cementitious composite flexure test at 7 days and the corresponding failure mechanism: Cm is the stress that the cementitious matrix can withstand before failure and Cr is the stress increase that the resin can resist before reaching point of failure. Cross section A-A depicts main crack perpendicular to the reinforcement at midpoint. M ¼ plain Mortar, SE ¼ Epoxy coated Sisal (S) reinforcing M, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S.
Table 5 Mean flexural strength of the various composites (MPa).a Days
M
S
SE
SV
SP
SPU
St
7 14 28
4.03 (5.28) 5.78 (6.96) 6.03 (8.36)
e e 6.09 (1.74)
12.39 (9.09) 18.66 (18.23) 16.47 (9.37)
10.88 (21.02) 17.9 (8.43) 18.39 (7.33)
9.5 (41.43) 14.48 (23.65) 14.51 (6.44)
14.35 (9.57) 19.68 (2.62) 19.8 (9.17)
e e 21.18 (4.47)
M ¼ Mortar, S ¼ Sisal, SE ¼ Epoxy coated S, SV ¼ Vinylester coated S, SP ¼ Polyester coated S, SPU ¼ Polyurethane coated S, St ¼ Steel fibre. a Value in ( ) is coefficient of variance (CV) in %.
It is apparent that the flexural strength varies between various combinations (Table 5). The variation of strength with curing times is because the cementitious matrix takes time to set. A scrutiny of the results shows that the flexural strength of SE- and PU-matrix composites is higher than that of SP- and SV-matrix composites, especially in the early stages of curing; however, this may not be accurate due to the high coefficient of variance. This is in agreement with different quality of the interface, as discussed in section 3.2, which is probably due to the compatibility and level of reaction between the polymer and natural fibres. This is also the case for the functional groups existed on the natural fibre and polymer surface, as discussed in section 3.1. All natural rebar composites are stronger than cement-based composites reinforced with sisal fibres (S), which is found to be the weakest (Table 5 and Fig. 3c). This highlights the significant benefit of using resin coating for natural fibre rebar as a dramatic increase in performance is witnessed. As seen, the maximum strength of SPU or SV reinforced composites is more than 3 times that of sisal fibre reinforced composites (Table 5 and Fig. 3c). Although the steel reinforced composite is the strongest, the strength at 28 days for the steel composites is only about 7% higher than that of SPU composites. This suggests that the widespread use of steel reinforcement in contemporary society may be reduced in some structural applications by potential reinforcements of comparable strength such as polyurethane (PU) coated natural fibre.
composites have been thoroughly investigated to determine the bonding and failure mechanisms. Some specific conclusions can be drawn as follows:
4. Conclusions
References
Natural fibre rebars have been developed and their use as reinforcement in structural concrete members has been tested in this paper. The behaviour and interface of the developed
1) A less compact interface was found when bare natural fibre rebar composite was used in comparison to coated natural fibre rebar composites, resulting in lower mechanical properties; 2) When the natural fibre is coated with certain resins such as polyurethane, the composite could attain mechanical (e.g. flexure) strength rivalling that of conventional steel reinforced composites; 3) All coatings on sisal rebars resulted in significant improvement in bonding interface and mechanical properties of the composites, albeit to varying degrees. It appears that the structure and quality of the resin coating influenced the strength of the composites depending on how well it protected the fibres from moisture; 4) The structural members with coated natural fibre rebars went through three stages of failure process with two clear dips in the stress-strain plots: the first dip occurred at similar flexure strains of approximately 10% regardless of the composite type, while the second dip originates from the mechanical failure of the resin coating with flexure strains reaching up to about 30% depending on the type of resin coating applied. Load capacity after the second dip depended on the architecture of the rebars.
[1] R.S. Blackburn, Biodegradable and Sustainable Fibres, Woodhead Publishing Limited, Cambridge, 2005. [2] A. Bentur, S. Mindess, Fibre Reinforced Cementitious Composites, second ed., Taylor & Francis, London and New York, 2007.
52
H. Ahmad, M. Fan / Cement and Concrete Composites 87 (2018) 44e52
[3] M. Fan, Characterization and performance of elementary hemp fibres: factors influencing tensile strength, BioResources 5 (2010) 2307e2322. [4] T. Gurunathan, S. Mohanty, S.K. Nayak, A review of the recent developments in biocomposites based on natural fibres and their application perspectives, Compos. Part A Appl. Sci. Manuf. 77 (2015) 1e25. [5] A. Mohanty, M. Misra, L. Drzal, Natural Fibers, Biopolymers, and Biocomposites, CRC Press, Boca Raton, 2005. €m, L.A. Berglund, R.D.T. Filho, Morphology and me[6] K. Oksman, L. Wallstro chanical properties of unidirectional sisal- epoxy composites, J. Appl. Polym. Sci. 84 (2002) 2358e2365. [7] P. Wambua, J. Ivens, I. Verpoest, Natural fibres: can they replace glass in fibre reinforced plastics? Compos. Sci. Technol. 63 (2003) 1259e1264. [8] X. Zhou, S.H. Ghaffar, W. Dong, O. Oladiran, M. Fan, Fracture and impact properties of short discrete jute fibre-reinforced cementitious composites, Mater. Des. 49 (2013) 35e47. [9] Food and Agricultural Commodities Production/Countries by Commodity, FAOSTAT, 2013 (fao.org). [10] Y. Li, Y. Mai, L. Ye, Sisal fibre and its composites: a review of recent developments, Compos. Sci. Technol. 60 (2000) 2037e2055. [11] A. Belaadi, A. Bezazi, M. Bourchak, F. Scarpa, Tensile static and fatigue behaviour of sisal fibres, Mater. Des. 46 (2013) 76e83. [12] C. Stevens, J. Müssig, Industrial Applications of Natural Fibres: Structure, Properties and Technical Applications, John Wiley & Sons, Bremen, 2010. [13] O. Onuaguluchi, N. Banthia, Plant-based natural fibre reinforced cement composites: a review, Cem. Concr. Compos 68 (2016) 96e108. [14] F. Gentles, Characterisation of Environmentally Friendly Fibres for Composites, University of Strathclyde, 2015. [15] F. Matthews, R. Rawlings, Composite Materials: Engineering and Science, Woodhead Publishing Limited, London, 1999. [16] BS EN 197-1:2011, Cement Part 1: Composition, Specifications and Conformity Criteria for Common Cements, BSI Stand., 2011. [17] J. Gassan, A. Bledzki, Possibilities for improving the mechanical properties of jute/epoxy composites by alkali treatment of fibres, Compos. Sci. Technol. 59 (1999) 1303e1309. [18] M. Rong, M. Zhang, Y. Liu, G. Yang, H. Zeng, The effect of fiber treatment on the mechanical properties of unidirectional sisal-reinforced epoxy composites, Compos. Sci. Technol. 61 (2001) 1437e1447. [19] T. Doan, H. Brodowsky, E. M€ ader, Jute fibre/epoxy composites: surface
[20] [21]
[22] [23] [24] [25]
[26]
[27] [28]
[29]
[30] [31]
[32] [33] [34]
properties and interfacial adhesion, Compos. Sci. Technol. 72 (2012) 1160e1166. W. Liu, T. Xie, R. Qiu, M. Fan, N-methylol acrylamide grafting bamboo fibers and their composites, Compos. Sci. Technol. 117 (2015) 100e106. K.L. Pickering, M.G. Aruan Efendy, T.M. Le, A review of recent developments in natural fibre composites and their mechanical performance, Compos. Part A Appl. Sci. Manuf. 83 (2015) 98e112. M. Ali, X. Li, N. Chouw, Experimental investigations on bond strength between coconut fibre and concrete, Mater. Des. 44 (2013) 596e605. BS EN 196-1:2005, Methods of Testing Cement d Part 1: Determination of Strength, BSI Stand, 2005. S. Sen, S. Patil, D.S. Argyropoulos, Thermal properties of lignin in copolymers, blends, and composites: a review, Green Chem. 17 (2015). K.L. Pickering, M.G.A. Efendy, T.M. Le, A review of recent developments in natural fibre composites and their mechanical performance, Compos. Part A Appl. Sci. Manuf. 83 (2016) 98e112. M.E.A. Fidelis, R.D. Toledo Filho, F. de A. Silva, V. Mechtcherine, M. Butler, S. Hempel, The effect of accelerated aging on the interface of jute textile reinforced concrete, Cem. Concr. Compos 74 (2016) 7e15. J. Hwang, Effects of Cement Treatment on the One-dimensional Consolidation Behavior of a Highly Organic Soil, Purdue University, 2006. F.D.A. Silva, B. Mobasher, C. Soranakom, R.D.T. Filho, Effect of fiber shape and morphology on interfacial bond and cracking behaviors of sisal fiber cement based composites, Cem. Concr. Compos 33 (2011) 814e823. a, P.R.L. Lima, Finite strip K. De Jesus Nagahama, A. De Souza Matos Gade modeling of cementitious laminates reinforced with sisal fibers, Cem. Concr. Compos 63 (2015) 8e16. F. de A. Silva, B. Mobasher, R.D.T. Filho, Cracking mechanisms in durable sisal fiber reinforced cement composites, Cem. Concr. Compos 31 (2009) 721e730. J.D.A.M. Filho, F.D.A. Silva, R.D. Toledo Filho, Degradation kinetics and aging mechanisms on sisal fiber cement composite systems, Cem. Concr. Compos 40 (2013) 30e39. BS EN ISO 178:2010þA1:2013, Plastics d Determination of Flexural Properties, 2011. G. Socrates, Infrared and Raman Characteristic Group Frequencies: Tables and Charts, John Wiley & Sons, London, 2004. M. Hesse, H. Meier, B. Zeeh, Spektroskopische Methoden in der organischen Chemie, eighth ed., Thieme, 2014.