Materials and Structures DOI 10.1617/s11527-006-9171-3
O R I G I N A L A RT I C L E
Properties of fabric–cement composites made by pultrusion A. Peled Æ B. Mobasher
Received: 9 May 2005 / Accepted: 10 July 2005 RILEM 2006
Abstract Practical manufacturing and use of thin cement-based elements composites require an industrial cost-effective production process in addition to proper reinforcement materials to improve the tensile and flexural performance. Reinforcement by means of fabric materials is an alternative to the use of short fibers. The objective of this study was to investigate use of pultrusion technique as a cost-effective method for the production of thin-sheet fabric-reinforced cement composites. Woven fabrics made from low modulus polypropylene (PP) and glass meshes were used to produce the pultruded cement composites. The influence of fabric type, PP and glass, processing method, pultrusion vs. cast and cement-based matrix modification were examined. Tensile and pullout tests as well as Scanning Electron Microscopy (SEM) observations were used to examine the mechanical, bonding and microstructure properties of the different composites. The rheology of the mix was correlated with the mechanical behavior of the pultruded composites. The tensile behavior of the pultruded A. Peled (&) Structural Engineering Department, Ben Gurion University, Beer Sheva, Israel e-mail:
[email protected] B. Mobasher Department of Civil and Environmental Engineering, Arizona State University, Tempe, AZ, US
fabric–cement components exhibited strain hardening behavior. The best performance was achieved for the PP pultruded composites. Keywords Cement composite Æ Textile Æ Processing Æ Pultrusion Æ Bonding Æ Cracking
1 Introduction The past decade has seen an increased use of pre-fabricated cement-bonded fiberboard around the world. Such elements are used for wall panels, exterior siding, pressure pipes, and roofing and flooring tiles. Use of reinforcement in these elements is essential in order to improve the tensile and flexural performance. The reinforcements can be either as short fibers or continuous reinforcements in a fabric form. Recently, several researchers have shown very promising results with thin sheet cement-based products reinforced with hand lay-up of fabrics [1–3]. In addition to ease of manufacturing, fabrics provide benefits such as excellent anchorage and bond development [4]. It has been shown that cement-based composite products with fabrics are much stronger than similar composite reinforced with straight continuous yarn (not in a fabric form), both made from the same low modulus polyethylene yarns [5–7]. The improved performance of fabric reinforced composites is due to the enhanced
Materials and Structures
matrix-reinforcement bond and mechanical anchorage caused by the geometry of the continuous yarns within the fabric [8, 9]. These enhancements are more profound in cementbased composites as compared to polymer composites where straight yarns lead to optimal results while non-uniform yarns reduce the reinforcing effectiveness [8]. Practical use of fabric–cement composites requires a cost-effective industrial production process such as the pultrusion method. This method is based on a relatively simple set-up using low cost equipment and has been used to produce cement composites with continuous filaments (filament winding technique) exhibiting significantly improved performance. Cement composites containing 5% (AR) unidirectional continuous glass fibers produced by pultrusion achieved tensile strength of 50 MPa [10], compared to an average tensile strength of about 6–10 MPa of conventional Glass Fiber Reinforced Cement (GFRC) composites. It is noted that the production method can have substantial impact on the properties of the final product [11]. The objective of this study was to document the mechanical response of high performance thinsheet fabric-reinforced cement composites manufactured by the pultrusion technique. This paper compares the effects of two different processing methods of casting and pultrusion. The variations in material formulations included rheology modifications using superplascticizers, silica fume, and fly ash. Bonded glass mesh fabrics and warp knitted weft insertion polypropylene (PP) fabrics were used to produce the various composites in this study. Tensile and pullout tests were used to characterize the mechanical performance and the bond between the fabric and the cement matrix. A microstructural analysis was also conducted and correlated with the mechanical test data.
2 Experimental program
yarns (warp and weft) are glued together at the junction points. In weft insertion warp knitted fabric the yarns in the warp direction are knitted into stitches to assemble together straight yarns, which are in the weft direction. The straight yarns (weft yarns) are the reinforcing yarns in the composite. The bonded fabric was made from multifilament AR with 2 yarns per cm, in both directions of the fabric. The weft insertion knitted fabric was made from multifilament PP with 8 yarns per cm in the reinforcing direction (weft yarns) and 0.8 stitches per cm in the perpendicular direction (warp yarns). The AR glass fibers were with tensile strength of 1270–2450 MPa, elasticity modulus of 78,000 MPa and filament diameter of 13.5 lm. The PP fibers had a tensile strength of 500 MPa, modulus of elasticity of 6900 MPa, bundle diameter of 400 lm, and filament diameter of 40 lm, giving 100 filaments in one bundle. 2.2 Cement paste mixtures In order to develop a mixture with optimal penetrability within the fabric openings tertiary blends of cement–fly ash–silica fume in addition to superplasticizer (SP) were used. The mix designs at the fresh stage of the matrices used in this study are presented in Table 1. Only the solid ingredients are presented by volume fraction, while in all mixtures the water/binder ratio by weight was 0.37. The rheology properties of the fresh mixtures were measured by shear rheometery 10 min after mixing. This delay represented the average time required to produce the pultruded sheet. 2.3 Composite preparation Composite specimens were prepared by the pultrusion process or by lay-up of fabrics (casting) in a cement matrix. The effect of the processing methods used was determined by studying the overall properties of the composite in tension as well as the bond behavior by pullout.
2.1 Fabric types 2.3.1 Specimens for tension tests Two types of fabrics were used for this study: bonded fabric and warp knitted weft insertion fabric. In bonded fabrics a perpendicular set of
Pultrusion process. In the pultrusion process, as shown in Fig. 1, fabrics were passed through a
Materials and Structures Table 1 Ingredients of the pultruded mixtures Volume fraction, % Batch #
#1
#2
#3
#4
Cement Silica fume Fly ash Superplasticizer
42 5 – 0.1
42 5 – 0.2
19 5 24 0.05
37 10 – 0.2
slurry infiltration chamber, and then pulled through a set of rollers to squeeze the paste in between the fabric openings while removing excess paste. The fabric–cement composite laminate sheets were then formed on a plate shaped mandrel, resulting in samples with width 25 mm, length 250 mm, and thickness of 9–10 mm. Each cement board was made with six layers of fabrics, resulting in a reinforcement content of about 6 and 9% by volume of AR glass fabric and PP fabric, respectively. The reinforcing yarns in the composite of each fabric are those along the reinforcing direction and the reinforcement content is taking into account only those yarns in both cases. Note that the volume fractions reported here were calculated based on the bundle diameter, assuming no penetration of the cement matrix in between the filaments of the bundle. After forming the samples, pressure was applied on top of the fabric–cement laminates to improve penetration of the matrix in between the opening of the fabrics. A constant pressure of 15 kPa due to a 900 N load applied on the surface of the fabric–cement sheet was used for all specimens. Most of this pressure was reduced within 1 h after
Fig. 1 Set-up of the pultrusion process
the pultrusion process to a level corresponding to 100 N load (1.7 kPa). This pressure was maintained up to 24 h from the pultrusion process. Casting process. In this process, specimens were prepared by hand lay-up of fabric layers in the cement matrix. Similar to the pultruded specimens, a six layer fabric laminate of dimensions 250 · 25 · 10–13 mm was prepared that provided reinforcing yarns by volume of 3.6 and 5.3% of the glass and PP fabric, respectively. Note that due to the increased matrix content in this case thicker specimens were prepared, resulting in a lower volume content of reinforcing yarns as compared to the pultruded specimens. To achieve a better penetration of the cement matrix in between the openings of the fabrics, vibration was applied to the samples during preparation. 2.3.2 Specimens for pullout tests The specimens were prepared by hand lay up of single layer of fabric sandwiched between cement paste layers. In order to examine the processing effect of pultrusion versus casting on the bond between the fabric and the cement matrix, two sets of specimens were prepared for each system. In the first set, ‘‘clean’’ fabrics without cement were embedded in the cement matrix. In the second set, the fabrics were first impregnated in the cement bath using the pultrusion process; then the impregnated fabrics were embedded in the cement matrix. The embedded length of the fabric was equal to the length of the specimen: 12.7 mm long for the AR glass fabric, 12.7 and 7.6 mm long for the PP fabric. Pullout test specimens consisted of two parts, an embedded fabric section and a plain fabric section. The plain fabric section containing 8 yarns in the longitudinal direction had a free length of 25 mm between the gripped end and the embedded region. The free length was necessary for handling purposes and ensuring that the load could be distributed uniformly to all the yarns. All the specimens, cast and pultruded for tensile and pullout tests, were demolded after 1 day of moist curing. The specimens were cured for 3 days in 80C, 100% Relative Humidity (RH) and then stored in room environment until testing
Materials and Structures
in tension or pullout, at 7 days after the pultrusion process or casting. Five replicate samples of each system were prepared for testing. 2.4 Mechanical testing Direct tensile tests of the pultruded and cast laminates along the machine direction were performed on an MTS testing machine under closed loop controlled conditions using an elongation rate of 0.5 mm/min. Typical stress–strain curves representing the tensile behavior of each series of individual composites were chosen for comparison. Formation of distributed cracks throughout the tensile testing was documented. Pullout tests were carried out in an Instron testing machine at a crosshead rate of 0.25 mm/s Load–slip curves were recorded and the test was conducted until embedded fabric was completely pulled out. The maximum shear bond strength, smax, per unit external bundle surface was calculated assuming the yarn as a single reinforcing cylindrical element with a diameter equal to the effective diameter of the material in the bundle. Fragments of specimens obtained after tensile tests were dried at 60C and gold-coated for observation using Scanning Electron Microscopy (SEM) and correlated with the mechanical performance. 2.5 Crack spacing measurements Brittle matrix composites containing a high volume fraction of fibers with ultimate strain capacity higher than the matrix, form a series of parallel microcracking when subjected to uniform tensile loading. A potential way to document the nature of cracking is to use an automated procedure based on image processing to detect crack formation as a function of applied global strain. In addition, one can document the rate of formation of new cracks as a function of applied strain. Crack density can be used to ascertain both the degree of damage and the bond strength between the fibers and matrix. As the test specimens are subjected to uniaxial tensile test, a distributed cracking pattern forms throughout the loading cycle which was recorded continuously. Photographs of the specimen at 15 s
time intervals were taken using a digital frame grabber. An initial image of the specimen before any crack has started to form was selected as the original state. The procedure for crack determination was based on the assumption that changes in the intensity of the pixels within a region indicate the existence of a crack. Each image was sharpened using standard routines such as Laplacian filters, and subjected to segmentation. This is a process to separate the crack from the rest of the image by specifying threshold intensity for selection of pixels. All the pixels below certain intensity were designated as a crack. An image of specimen containing newly formed cracks is superposed on the initial image and cracks are traced. This process is repeated between each subsequent image and the results were added sequentially. Digital image processing toolbox of MATLAB was used and an algorithm was developed to measure the spacing between the traced cracks. This approach is described in detail in an earlier work [7]. The crack spacing is initially measured in pixels, and converted to length measures using conventional calibration of the image. The distribution of crack spacing measures at various strain levels can be used to better understand at the stages of cracking. For the low modulus fabric (PP) the increases in a single crack width was recorded (the first visible crack) using a procedure similar to the crack spacing measurements. This procedure was used to compare the two PP fabric systems, pultruded and cast.
3 Results and discussion 3.1 Effects of fabric type and crack spacing The tensile responses of pultruded composites with AR glass and PP fabrics are presented in Fig. 2. More ductile behavior is seen for the PP fabric system compared to the AR glass composite, with greater tensile strength at high strains. At lower strain values similar behavior is observed for both fabric systems, despite the greater modulus of elasticity of the glass fabric (78 and 6 GPa for the glass and the PP, respectively). Note that the initial modulus is similar for
Materials and Structures
both fabric composite systems and mainly represents the modulus of the cement matrix; the influence of the fabric is obvious at the stage where cracks initiate. Parameters of crack spacing as a function of applied strain are correlated with the stress–strain plot and also shown in Fig. 2 for both fabrics. Note that using the image analysis approach significant number of measurements are collected from each position on the curve. The PP composite shows much smaller crack spacing compared to the glass fabric composites. This indicates the greater ability of the PP fabric in crack distribution, with ultimate crack spacing of almost 1/3 that of AR glass fabric. The saturation crack spacing is about 18 mm for the glass composite and 5 mm for the PP composite. Such behavior can be related to improve bonding and smaller size of the fabric opening in the PP system to be discussed later. The figure shows a general decrease in the crack spacing during loading in both cases until the curves reach a steady state. If no new cracks are forming, it is therefore reasonable to assume that additional imposed strain resulted in widening of the existing cracks. At this stage, where crack spacing saturation is first observed, the opening of the main crack (first visible one) is only 100 lm for the low modulus PP composite (about 2% strain level). At failure of the glass composite (at about strain level of 4%) the crack width of the PP is less than 200 lm. Distinct regions of the stress–strain response were selected and the crack spacing distributions 60
35
Stress, MPa
25
40
20 15 20
10
Crack spacing, mm
AR Glass Fabric PP Fabric
30
5 0
0
0.02
0.04
0.06
0 0.08
Strain, mm/mm
Fig. 2 Comparison of tensile stresses of AR glass and PP fabric systems as well as crack spacing as a function of applied strain, both made by the pultrusion method
in these ranges were evaluated by looking at the range and distribution of the data (Fig. 3). The distribution of the measurements conducted at each strain range represent the crack spacing measured from approximately 200 observations per sample per image. The cumulative distribution functions shown in Fig. 3c, d represent the distribution of the crack spacing for the three different ranges in the strain history (Fig. 3a, b). Note that initially, the cracks spacing is large, but as straining of the sample proceeds, the crack spacing becomes more uniform and the standard deviation decreases. These curves support the general observations that during the initial loading stage (e = 0.015, for the AR glass), the distribution is varying in two distinct ranges. Within the first two strain ranges selected, the spacing between any two cracks is systematically reduced due to new crack formation in between them. As the strain is increased, we observe that the crack distribution becomes homogenized and more uniform such that at strain levels of 3.87% (Fig. 3c), one can see that 80% of the cracks have a spacing of about 10 mm. In the PP specimens the same homogenization effect is taking place (Fig. 3d) while the final crack spacing numbers are much lower (5 mm) than the samples with glass fabrics (10 mm). 3.2 Effect of processing methods 3.2.1 Tensile performance Figure 4 presents the tensile behavior of AR glass fabric produced with the pultrusion method. Results are compared with the conventional GFRC containing short glass fibers and produced by the spray process. The benefit of using the pultruded fabric laminates is clearly observed. Compared with the conventional GFRC, superior tensile behavior in both strength and toughness is seen for the pultruded specimen. This improvement in tensile behavior is about 1.5 folds, indicating the advantage of using continuous fibers in the form of fabrics with the pultrusion process. The short glass fibers were ‘‘clean’’ with a very thin sizing material which allowed a well dispersed distribution of fibers with in the cement matrix as opposed to the coated fabric with its yarns
(a) 35
(c)
30 25
Stress, MPa
Fig. 3 (a) and (b) tensile response of AR glass fabric (with fly ash), and PP fabric, with selected three strain ranges, (c) and (d) the cumulative crack spacing distributions as a function of crack spacing measured at three strain ranges for AR glass fabric and PP fabric
Cumulative Distribution Function
Materials and Structures
Zone 3
20
Zone 2
15 Zone 1
10 AR-Glass Fabric
5 0
0
0.02
0.04
0.06
1
0.8
Zone 2 ε = 0.0273
0.6
Zone 1 ε = 0.015
0.4 AR-Glass Fabric
0.2
0
0.08
Zone 3 ε = 0.0387
0
Strain, mm/mm
(b) 35 30
Cumulative Distribuition Function
(d) Zone 3
Stress, MPa
25
Zone 2 20 15 Zone 1 10
PP Fabric
5 0
0
0.02
0.04
0.06
0.08
1
60
1 0
0.8
20
Zone 2 Zone 3
40
60 0.8
0.6
0.6
Zone 1 0.4
0.4
0.2 0
PP Fabric
0
20
2
0
40
60
Crack Spacing, mm
Strain, mm/mm
Fig. 6. Improvement in strength is observed for the pultruded composite with more brittle behavior. Note that this improvement by the pultrusion process is much smaller than that observed in the case of the PP fabric system (Fig. 5). Figure 6 also shows the crack spacing as
20 Tensile Stress, MPa
remaining relatively un-dispersed. This may lead to the greater composite stiffness after cracking of the short GFRC as compared with the fabric– cement composites. Fracture of the short fibers could however occur more easily during testing, leading to more brittle behavior. No fracture of the reinforcing glass yarns of the coated fabric was observed during testing. Note that the GFRC are commercial samples and used for comparison with the fabric–composite prepared for this study. The tensile behavior of pultruded and cast specimens using the PP fabrics is presented in Fig. 5. The uniform tensile behavior of the pultrusion specimens compares to much less consistent tensile behavior of the cast specimens is observed. The post-cracking stiffness of the pultruded specimens are much higher than the cast systems suggesting an improved bond. This indicates the advantage of the pultrusion process in assuring uniform production of fabric–cement laminates. Similar trend was also observed with the glass system. Comparison of the two processing methods of cast and pultrusion for glass fabrics is presented in
20 40 Crack Spacing, mm
AR Glass Fabric vf=6%
16 GFRC Vf =5%
12 8 4 0
0
0.01 0.02 Strain, mm/mm
0.03
Fig. 4 Tensile behavior of AR glass fabric produced with the pultrusion method compares with GFRC produced by the spray process containing short glass fibers
Materials and Structures
Pultrusion Cast
12 8 4 0
0
0.02 0.04 0.06 Strain, mm/mm
0.08
Fig. 5 Comparison of pultrusion vs. cast on the tensile behavior of specimens with PP fabrics
a function of the strain. Note that as the strain on the sample is increased both the average crack spacing and its standard deviation reduce significantly. Uniform crack spacing is obtained through the later stages of loading, indicating that at these loading stages, only crack opening is taking place as opposed to formation of new cracks. Smaller crack spacing of about 10 mm, is observed for the pultruded system as compared to measurements of 20 mm for cast specimen. Similar trend of
80 AR Glass Fabric
Stress, MPa
16
Cast Pultruded
60
12 40 8 20
4 0
0
0.01 0.02 0.03 Strain, mm/mm
In order to better understand the influence of the pultrusion process, the bond between the fabrics and the cement paste matrix was studied using pullout tests. Impregnated fabrics using the pultrusion process (pultruded) are compared to nonimpregnated glass and the PP fabrics (cast) are shown in Fig. 8. Table 2 presents the average bond strengths for different fabrics and processing methods and slips at maximum stress. A marginal influence of the production process on pullout resistance and bond strength is observed with the glass composites (Fig. 8a and Table 2). However, similar observations on the
40
Crack spacing, mm
20
3.2.2 Pullout behavior and bonding
0 0.04
Fig. 6 Tensile response of pultruded and cast composites with AR glass fabric
1.2 1
Tensile Stress, MPa
Stress, MPa
16
30 0.8 Pultruded Cast
20
0.6 0.4
10
Crack Width, mm
20
denser crack pattern with the pultruded composites was also obtained with the PP system, suggesting better bonding in the case of the pultruded composites. Crack width as a function of applied strain was tested on the low modulus PP specimens (Fig. 7). The width of the crack through the entire loading is significantly larger for the cast composite than the pultruded one. At a strain value of about 1.5%, the cast crack width is 0.4 mm, but only 0.1 mm in the pultruded specimen. At a higher strain, 6%, this difference is even more evident, showing with crack width values of 1 and 0.2 mm for cast and pultruded specimens, respectively. These observations confirm the improved bonding of the PP fabric with the pultrusion process.
0.2 0
0
0.03
0.06
0.09
0 0.12
Strain, mm/mm Fig. 7 Effect of crack width during loading of the PP fabric composite, for cast and pultruded specimens
Materials and Structures
PP systems are quite noticeable. The pullout resistance of the pultruded PP systems (impregnated fabrics) is much greater than those of the cast systems. This improvement in bonding is clearly pointed out in Table 2, as the bond strength of the pultruded PP is about two times as that of the cast fabric (for 7.6 embedded length). The improved bonding is correlated with the superior tensile behavior of the pultruded PP composites (Fig. 5). The bond strengths of the glass systems for both pultruded and cast fabrics are relatively poor compared with the PP systems, suggesting better bond of the PP fabric with the cement matrix (Table 2). This improved bonding of the PP fabric system is quite pronounced and can lead to the increased tensile performance of the pultruded PP composites (Fig. 2). Note that the slip at
(a) 300 Glass Fabric
Pullout Load, N
250
Pultruded Cast
200 150 100 50 0
0
3
6 Slip, mm
9
12
(b) 300 PP Fabric
Pultruded Cast
Pullout Load, N
250
maximum shear stress is much smaller for the glass systems as compared with the PP. This may imply that the width of the main crack of the glass system at strain level of 2% is smaller than the PP which at this strain level exhibits a crack opening of 0.1 mm. 3.2.3 Microstructure characteristics Figure 9 presents SEM micrographs of glass fabric and PP fabric embedded in the cement matrix for the two processing methods, pultrusion and casting. Penetration of the cement matrix in between the reinforcing filaments of the bundle is well developed for the PP pultruded fabric as shown in Fig. 9a. Impregnation of cement paste is also seen in between the stitches of the pultruded composite. The matrix fills the spaces between the filaments of the stitches as well as in between the loops. Much poorer penetration is observed with the PP cast composite, leaving empty spaces in between the filaments of the bundle and the loops are relatively empty (Fig. 9b). No penetration is possible for the glass fabric as the bundles are coated with sizing (Fig. 9c). Therefore, no significant difference in microstructure was observed for the pultruded and cast glass composites. These observations indicate that the PP fabric is more affected by the processing methods: better bonding leads to improved performance of the composite and is obtainable with the pultruded PP composite compared to cast composites (Figs. 8b and 5). However, no such significant difference between the two processing methods is seen for the glass system for the overall mechanical behavior (Fig. 6) as well as bonding (Fig. 8a). These differences in behavior between
200 150
Table 2 Bond strength of the different fabrics and processing
100
Specimen type
50 0
Embedded Bond strength, Slip at length, mm MPa maximum shear stress, mm Pultruded Cast Pultruded Cast
0
3
6 Slip, mm
9
12
Fig. 8 Pullout load–slip response of the different fabrics: (a) glass and (b) PP
PP fabric 7.6 PP fabric 12.7 Glass fabric 12.7
2.91 2.14 0.57
1.55 1.76 1.44 3.10 0.72 0.69
1.38 1.72 0.74
Materials and Structures
the glass and the PP systems are associated with the differences in the structure of the fabrics and yarns which make up the fabrics. The PP fabric is a knit type fabric which is made up of bundles and connected at the joint points by stitches, whereas the AR glass fabric is a bonded type fabric and the yarns were impregnated in sizing prior to the
production of the fabric. When the bundled PP knit fabric is used to produce composite by the cast process, the penetration of the paste into the spaces between the filaments in the bundles is relatively low. This is due to the presence of the bulky stitches themselves, as well as the tightening effect of the stitches, which strongly hold the filaments in the bundle and prevents them from being opened. Therefore, poor penetration of the matrix in between the filaments takes place, preventing the reinforcement potential of the individual filaments to be materialized. However, when this bundled knit fabric is passed through a cement bath during the pultrusion process, the intensive shear stresses during impregnation help fill the spaces between the filaments of the bundled yarns as well as the loops of the stitches. This leads to improved bonding and enhanced mechanical performance (Figs. 8b and 5). On the other hand, with the glass fabric the bundled yarns were impregnated with epoxy prior to fabric production, gluing the filaments together and filling the spaces in between them; thus, there is no room to force the matrix paste in between the filaments and the pultrusion does not offer any advantage. This may explain the improved tensile performance of the pultruded PP as compared to the glass system (Fig. 2). It should be remembered that for the pultruded specimens a pressure was applied on top of the composite after casting in order to improve the bond as compared to the cast composite [11]. This may explain the greater tensile strength of the glass pultruded composites as compared with the cast composite (Fig. 6). However, since no pressure was applied on top of the pullout specimens, improvements for the pultruded samples were not observed (Fig. 8a). 3.3 Matrix modification
Fig. 9 SEM observations of (a) PP pultruded composite (b) PP cast composite and (c) glass pultruded composite
Based on the above discussion, it is clear that the fabric impregnation by the cement paste is an important factor during the pultrusion process. In order to better recognize these rheology effects in pultruded systems addition of SP, silica fume and fly ash were evaluated (Table 1). Figure 10 presents the effect of the viscosity of the various fresh cement matrix (shear stress) versus the
Shear Stress, D/cm^2
Materials and Structures 8000
10% Silica fume
6000 No fly ash
4000 2000
With fly ash
0 10
13
5% Silica fume 16 19 22
25
Tensile Strength, MPa
Fig. 10 Effect of fly ash and silica fume addition on the tensile behavior of the fabric–cement composites
tensile strength of pultruded composites made with glass fabric. It can be clearly observed that the viscosity properties of the fresh cement matrix significantly affects the tensile strength of the glass fabric composites. Stiffer matrices reduce the mechanical properties of the cement composite. The matrix with the addition of fly ash shows the lower shear stress (the most fluidity) and the best tensile strength of the composite. Better fluidity of the fresh matrix can lead to better penetration of the cement in between the fabric opening and improved mechanical properties as observed with the pultruded PP fabric system. 4 Conclusions The pultrusion technique for the production of fabric–cement products requires relatively simple set-up using low cost equipment while allowing good control of laminates alignment giving relatively smooth surface and uniform products. The pultrusion process was found to significantly improve the mechanical performance of the cement composites compared with cast composite when ‘‘clean’’ bundle yarn (not impregnated) are used to produce the fabrics. In such fabrics the pultrusion process enables the spaces between the filaments of the bundle to be impregnated with the paste, resulting in a much better bond and maximize the efficiency of the filaments, leading to improved tensile performance. In fabrics where the bundles of filaments are impenetrable due to the use of sizing, there is no advantage to the pultrusion process from a mechanical point of view, since there are no fine
spaces to force the matrix into. Thus, casting and pultrusion processes gave similar results. The mobilization of the filaments in the pultrusion process results in a strain hardening composite, even when the modulus of the yarn is relatively low. Therefore, improved mechanical performance of low modulus fabric such as the PP was observed as compared to fabric with much higher modulus of elasticity as the glass studied here. This can be used as a concept for mobilizing low modulus yarn fabrics to obtain high performance cementitious composites. The rheological properties of the fresh matrix found to significantly correlate with the mechanical performance. Increasing the fluidity of the matrix improves mechanical performance of the composite. Acknowledgements The authors would like to thank Nippon Electric Glass Co., Ltd., Kuraray America Inc., and Karl Mayer Ltd. for their cooperation for providing the fabrics used in this study. The National Science Foundation, Program 0324669-03, and the BSF (United States Israel Binational Science Foundation) are acknowledged for the financial support in this research.
References 1. Meyer C, Vilkner G (2003) Glass concrete thin sheets prestressed with aramid fiber mesh. In: Naaman AE, Reinhardt HW (eds) Proceeding of the 4th international RILEM workshop on high performance fiber reinforced cement composites (HPFRCC4). Ann Arbor, pp 325–336 2. Kruger M, Ozbolt J, Reinhardt HW (2003) A new 3D discrete bond model to study the influence of bond on structural performance of thin reinforced and prestressed concrete plates. In: Naaman AE, Reinhardt HW (eds) Proceeding of the 4th international RILEM workshop on high performance fiber reinforced cement composites (HPFRCC4). Ann Arbor, pp 49–63 3. Nishigaki T, Suzuki K, Matuhashi T, Sasaki H (1991) High strength continuous carbon fiber reinforced cement composite (CFRC). In: Brandt AM, Marshall IH (eds) Proceeding of the 3rd international symposium on brittle matrix composites. Elsevier Applied Science, Warsaw, Poland, pp 344–355 4. Swamy RN, Hussin MW (1990) Continuous woven polypropylene mat reinforcement cement composites for applications in building construction. In: Hamelin P, Verchery (eds) Textile composites in building construction, Part 1, pp 57–67 5. Peled A, Bentur A, Yankelevsky D (1999) Flexural performance of cementitious composites reinforced by woven fabrics. Materials in Civil Engineering (ASCE), 11(4): 325–330
Materials and Structures 6. Mobasher B (2003) Micromechanical modeling of Angle ply filament wound cement-based composites. ASCE, J Eng Mech 129(4):373–382 7. Mobasher B, Peled A, Pahilajani J (2006) Distributed cracking and stiffness degradation in fabric–cement composites. Mater Struct, in press 8. Bentur A, Peled A, Yankelevsky D (1997) Enhanced bonding of low modulus polymer fibers-cement matrix by means of crimped geometry. Cement Concrete Res 27:1099–1111
9. Peled A, Bentur A, Yankelevsky D (1998) Effects of woven fabrics geometry on the bonding performance of cementitious composites: mechanical performance. J Adv Cem Based Mater J 7:20–27 10. Mobasher B, Pivacek A, Haupt GJ (1997) cementbased cross-ply laminates. J Adv Cem Based Mater 6:144–152 11. Peled A, Mobasher B (2005) Pultruded fabric–cement composites. ACI Mater J 102(1):15–23
