EFFECTS OF WARP KNITTED FABRICS MADE FROM MULTIFILAMENT IN CEMENT-BASED COMPOSITES

Transcription

1 1 st International Conference Textile Reinforced Concrete (ICTRC) 23 EFFECTS OF WARP KNITTED FABRICS MADE FROM MULTIFILAMENT IN CEMENT-BASED COMPOSITES Zvi Cohen (a), Alva Peled (b), Yonatan Pasder (a), Andreas Roye (c) and (c)thomas Gries (c) (a) Material Engineering Department, Ben Gurion University, Beer Sheva, Israel (b) Structural Engineering Department, Ben Gurion University, Beer Sheva, Israel (c) Institut für Textiltechnik at RWTH Aachen University (ITA), Germany ABSTRACT: The objective of this study was to explore the influences of different characteristics of warp knitted fabrics made from multifilament yarns on the tensile properties of TRC elements and the bonding developed between the fabrics and the cement matrix. Several parameters such as loop size, bundle size (number of filaments), and fiber type (different raw materials: High Density Polyethylene, Polypropylene, AR glass and Aramid) were examined. All the composite elements were produced by the pultrusion technique. It was found that the loop size of the fabric as well as the bundle diameter significantly influence the performance of the cement composite. Fabrics made from small bundle diameter and large loop size exhibited the highest efficiency factor and developed the best bond strengths with the cement matrix. This was explained based on tighten effects induced by the stitches of the fabric and cement penetrability in between the filaments of the bundle. 1 INTRODUCTION Recently, a research was conducted on the use of textile fabrics as reinforcement for cementbased composite materials (TRC) [Ban04, Jes03, Rei03, Roy04]. The reinforcement of cement by textiles was to improve mechanical properties of elements that require improved tensile and flexural performances. Several researchers reported that fabrics could significantly improve the mechanical behavior of cement matrices. In addition to the improved strength, these fabric-reinforced cement composites exhibited a strain-hardening behavior even when the reinforcing yarns had a low modulus of elasticity. This was explained by an enhancement in bonding due to the mechanical anchoring provided by the nonlinear geometry of the individual yarns within the fabric, induced by the fabric structure [Pel98a, Pel99b, Pel00c]. Textile fabric reinforcements can be knitted, woven, non-woven, glued or plaited. These forms of fabric structure differ by the manufacturing process and several other parameters such as yarn s density, fineness, and number of filaments in the bundle that can influence the stability and the mechanical properties of the whole fabric. When the fabric is composed of multifilament yarns (bundles) the potential penetrability of the cement into the bundle spaces is lower than that of single bundles (not in a fabric form) due to tightening effects induced by the junction points of the fabric, which strongly hold the filaments of the bundle and prevents them from being opened. The penetrability of the matrix into the fabric and especially in between the filaments depends upon the nature of the fabric junctions and the resulting tightening effects, the structure of the fabric, the number of filaments in the bundle, and the production process of the composite. Preparation of cement composites with textile fabrics requires an appropriate process that will ensure a good penetrability of the cement and a proper bonding between the cement matrix

2 24 COHEN,PELED,PASDER,ROYE,GRIES: Effects of warp knitted fabrics made from multifilament and the fabric. Pultrusion is one technique that can provide such improvement [Pel00d, Pel03e, and Pel03f]. It helps the cement matrix to penetrate in between the filaments of the bundle and in between the opening of the fabric, in doing, so it generates composite elements that have superior mechanical properties than those achieved with ordinary casting. In this work weft insertion warp knitted fabrics made from multifilament yarns as reinforcements for cement composites were studied. Influences of parameters related to these knit fabrics such as loops size, bundles size (number of filaments), and fiber type such as Aramid, AR glass, High Density Polyethylene (HDPE), and polypropylene (PP) on tensile performances of cement composites were examined. A correlation between mechanical performance of the fabrics themselves (not in a matrix) and the fabric-cement composites was examined. All composites in this study were prepared by the Pultrusion process. Bond strength and adherence of these systems were also studied. 2 EXPERIMENTAL 2.1 Fabrics and yarns Two dimensional (2D) types of warp knitted fabrics were specially prepared for this study on a two needle bar Raschel machine (Karl Mayer) at Institut für Textiltechnik at RWTH Aachen University (ITA). In these knitted fabrics, straight yarns in the warp direction (lengthwise) are inserted in stitches (loops) and assembled together with straight yarns in the weft direction (crosswise). In both directions of the fabric (x and y) the yarns were straight, referred here as 0 for the straight warp yarns and 90 for the weft insertion yarns. A density of 6 stitches (0 yarns) per 2.4 mm (E6) was used during preparation of the fabrics, giving a grid net like fabrics to help cement penetrability. All yarns, warp and weft were in a multifilament form, non-twisted. All stitches in all fabrics were made of PP with 16.7 tex. The weft yarns in all fabrics were composed from Aramid with 167 tex. The reinforcing yarns were those in the warp direction (0 ) composed of different raw materials: HDPE, Aramid, AR glass and PP. Four different types of HDPE fabrics were prepared and studied: two with different bundle diameter (i.e., different tex, 138 and 88) in order to check and determine the influences involved with the size of the bundle. Tex is the fineness of the yarn (bundle) given by the weight, in grams, of 1000 m of yarn. At each bundle size two fabrics were prepared with different loop sizes, 2 mm and 4 mm. The loop size resulted in a change of the distance between the wefts (90 yarns), 4 mm and 8 mm, respectively. A Difference in the fabric opening (weft yarns density) can affect the cement penetrability and anchoring of the fabric within the cement matrix. The geometry and properties of the yarns and fabrics are presented in Table 2.1. The bundle diameter was calculated by measuring the weight of the yarns, 2 m long, and by that calculating the tex number of the yarns. With the correlation between tex number and the cross section area (Eq. 2.2), the diameter was calculated. mass length = cross-section area X fiber density (2.1)

3 1 st International Conference Textile Reinforced Concrete (ICTRC) 2 Table Properties and geometry of the different fabrics and yarns made up the fabrics along the reinforcing direction (warp, 0 0 ) Yarn type Yarn tex Bundle Fabric Fabric tensile Fabric g/km diameter, mm loop size strength, MPa # Aramid mm AR glass mm HDPE (Dyneema) mm 4 mm HDPE (Dyneema) mm 4 mm PP mm Composite preparation All composites in this study were prepared in the pultrusion process. As shown in Fig. 2.1, during pultrusion, fabric passes through a slurry infiltration chamber and then is pulled through a set of rollers to squeeze the paste between the fabric openings while simultaneously removing excess paste. The fabric-cement composite laminate sheets were then formed on a plate shaped mandrel, resulting in samples with 30 mm width, 300 mm length, and 8 mm thickness. Each specimen was composed of 3 layers of fabrics. In all cases, the matrix was cement paste made from 0.4 water/cement ratio. Fig Pultrusion process After forming the samples, a constant pressure of 0 N was applied on top of the fabric cement laminate to improve matrix penetration between the yarn and fabric openings. This pressure was maintained up to 24 hours. Then each of the laminated sheets was cut to elements of 20 mm length, 34 mm width and 8 mm thickness and cured at 100% Relative Humidity and room temperature for 2 days. After 2 days of curing, the specimens were placed at room environment for another 3 days until testing in tension. 2.3 Tensile Tests Fabrics: The tensile properties of the different fabrics were studied with a closed loop control direct tensile testing machine with a capacity of 10KN. The rate of cross head displacement was set at 10 mm/min. The fabrics tested were cut along each warp yarn giving a strip of fabric that contains single warp yarn (0 ) covered with the stitches and the junctions with the

4 26 COHEN,PELED,PASDER,ROYE,GRIES: Effects of warp knitted fabrics made from multifilament weft yarns (90 ), based on a method developed at ITA, RWTH Aachen University [Roy04]. Tensile strengths were calculated for all fabrics. Composites: The mechanical performances of the pultruded composite laminates were studied using closed loop control direct tensile tests performed on a testing machine with a capacity of 10KN. The rate of cross head displacement was set at 0. mm/min. Metal plates with the dimensions of 30x0x1 mm were glued onto the gripping edges of the specimen to minimize localized damage and provide a better load transfer from the grips. At least, six replicate samples were tested in each fabric and yarn category; the reported results reflect the average and standard deviation values. Typical stress-strain curves representing the tensile behavior of individual composites were compared. All tests were ended at failure. The post cracking tensile stresses were calculated for all systems at different strain levels: maximum strain ( max, at failure), 2% ( 2% ), and 1% ( 1% ). The tensile strength reinforcing efficiency, defined as the ratio between the post cracking tensile strength and the product of the yarn volume and its tensile strength (composite tensile strength/ (volume fraction of the fibers x fiber strength)). The efficiency factor was calculated to compare the different fabric systems with the different yarns and volume content. The bond between the fabric and the matrix was calculated based on the ACK model (Eq. 2.1), this model deals with the mechanism of stress transfer from the fiber to the matrix and therefore causing a multiple cracking in brittle matrixes, by the mean crack spacing obtained at composite failure as presented in Eq V V m f mur 2X ' (2.1) Where: X'= means spacing between the cracks in the specimen [mm]; V m, V f = volume fraction of matrix and reinforcing yarns, respectively; r = radius of the reinforcing yarn in the fabric [mm]; mu - yield/ultimate strength of the matrix without reinforcing (4 MPa for cement); = bond strength between the matrix and the fabric [MPa]. 3 RESULTS AND DISCUSSION 3.1 Fabrics Properties Table 3.1 presents the tensile properties of all tested fabrics. The Aramid fabric is the strongest followed by the AR glass and HDPE. The PP fabric shows the lowest strength value as expected. The HDPE with the yarn having a lower tex value of 88 performed better than a similar fabric with a higher tex value of 138, mainly with 4 mm loop size. Improvement in fabric strength is also observed when the loop size is increased from 2 mm to 4 mm, for both HDPE fabrics with 88 tex and 138 tex. The improvement in fabric strength by the loop size is relatively high for the low tex case of about 20%, and less significant of about 10% in the case of the high tex fabric with 138 tex. 3.2 Properties of the Composites Table 3.2 presents the tensile strengths and composites efficiency factors of all tested systems for several strain levels of the composites: at maximum strain (failure), 2% strain, and 1% strain. The table also includes the bond strength values calculated, based on the

5 1st International Conference Textile Reinforced Concrete (ICTRC) 27 spacing between the cracks at the end of testing by ACK model. Several parameters are compared: raw materials of reinforcing yarns generating the fabrics, bundle size (yarn fineness, tex), and size of loops connecting the weft and warp yarns. The yarns were not twisted to prevent an influence of other parameters on the results. 3.3 Influences of fabric types Fig. 3.1 presents the tensile behavior of different fabrics made from various raw materials: high performance Aramid, AR glass, and HDPE, and low performance PP (Specimens # 1, 2, 3, and 7, respectively, see Table 3.2). Composites made from Aramid and HDPE fabrics exhibit similar tensile behavior with relatively high strengths at failure and 28.7 MPa, respectively, and improved toughness (Fig 3.1 and Table 3.2, Specimens 1 and 3). This is for fabrics having 2 mm loop size and similar yarn tex 138 and 167, for HDPE and Aramid, respectively. The lowest tensile strength values of only 9.2 MPa is observed for the PP composite. These results correlate with the properties of the fabrics themselves (Table 3.1). Note that above about 2% strain level no new cracks were developed for the HDPE, Aramid and PP systems; suggesting that from this point, further increase in the strain is due to widening of the cracks and the loads are mainly carried by the fabric itself. The Glass fabric composite shows a relatively brittle behavior as compared with all the other composite systems with a tensile strength at failure higher than that of the PP composite but lower than those of the HDPE and Aramid composites at max. strain level. However, at low strain value of 1% the tensile stress is the highest for the AR glass composite. Suggesting that for low-strain applications the AR glass composite is the best choice. The reinforcing efficiency factors are relatively high for the Aramid and AR glass composites as compared with the HDPE system (Specimens 1, 2 and 3 in Table 3.2). This factor takes into consideration the properties of the fabrics and volume content of reinforcement at each system. The improvement in the efficiency factor of the Aramid and AR glass systems is more than 0% over that of the HDPE composite. The tensile results are correlated with the bond strength values of the different systems. The greatest bonding is observed for the Aramid,.29 MPa, followed by.13 MPa for the AR glass, 4.4 MPa for the HDPE and only 2.4 for the PP. Note that the Aramid, glass and HDPE fabrics (yarns) have similar tensile properties (Table 3.1), however, due to the chemical nature of the glass and Aramid, they can develop a good bonding with the cement matrix, as observed here. No chemical bonding is expected between the HDPE and the cement matrix due to the significant difference in their chemical nature (hydrophilic for the cement and hydrophobic for the HDPE), resulting in poorer bonding and relatively low efficiency factor as compared with the other high performance fabrics. This can explain the improved tensile properties of the glass composites (at failure) and the Aramid composite at relatively low strain (lower than 2%) as compared with the HDPE at similar strain level. The high bond values of the AR glass system can lead to the brittle behavior of this composite, as the fabric tends to fail rather than to pull out during tensile test, exhibiting the brittle nature of this fabric. The HDPE and Aramid fabrics (yarns) are more ductile and therefore, a failure of the composites occurs at a much higher strain level due to the ductile nature of the reinforcing fabrics.

6 28 COHEN,PELED,PASDER,ROYE,GRIES: Effects of warp knitted fabrics made from multifilament Tensile Stress [MPa] mm loop size AR-Glass PP Aramid HDPE 0 0% 2% 4% 6% 8% Strain Fig. 3. 1: Stress-Strain responses of composites with different fabrics Table Tensile strengths of the composites and efficiency factors for several strain levels, including bond strengths Fabric - Specimen # V f % Bond Strength MPa At 1% strain At 2% strain At max strain Efficiency 1% Efficiency 2% Efficiency max factor MPa factor MPa factor MPa (0.3) (1.3) (8.83) (1.17) (0.93) (0.38) (1.28) (2.82) (0.88) (1.14) (3.62) (0.2) (0.87) (1.4) (1.07) (1.8) (3.88) (0.64) (0.83) (0.72) The numbers in brackets are the standard deviations 3.4 Influences of bundle size (tex) The influences of the bundle size (tex) is presented in Fig. 3.2 for a composite reinforced with HDPE fabrics made from two different bundle sizes of 88 tex and 138 tex, with a loop size of 4 mm (Specimens 4 and 6 in Table 3.2). It can clearly be observed that the composite system with 138 tex yarn performed much better than that with 88 tex at high strain level, above 2%. The improvement in composite stress values is about 7%. However, at low strain levels (below 2%), the tensile stresses are similar for both bundle systems, with 88 as well as with 138 tex. Note that above 2% strain level no new cracks were developed and an increase in the strain above this level is mainly due to crack widening while the loads are carried by the fabric. The understanding of these trends should consider several aspects: (i) properties of the

7 1st International Conference Textile Reinforced Concrete (ICTRC) 29 fabrics themselves, (ii) penetrability of cement particles in between the filaments of the bundle, i.e., bonding, and (iii) volume content of reinforcement. The low tex yarn is the one with the smaller diameter, 0.34 mm, providing a better possibility for the cement penetrability. The volume content of reinforcement is higher for the sizable bundle system (0.43 mm diameter), V f =2.7%, as compared with only 1.6% V f for the thinner bundle (0.34 mm diameter). The high volume content can explain the increased properties of the composite with the large bundle (Specimen 4 in Table 3.2) at high strain level, as the loads are mainly carried by the fabric itself. Note that the composite stresses were calculated based on the cross section of the entire composite also at high strain levels. However, at low strain values where progressive cracking is taking place and the loads are carried by the entire composite, cement penetrability in between the filaments of the bundle and the resulting bond strength can become dominant factors. The improved bond strength of the thin bundle composite, MPa, compared to that of the large bundle composite, 9.7 MPa, is clearly observed in Table 3.2 (Specimens 6 and 4, respectively) due to good cement penetrability. Therefore, at low strain value during multiple cracking, several opposite influences can affect the tensile behavior of the cement composite with the large diameter bundle: on the one hand, increase in composite stresses due to high volume content of reinforcement, but on the other hand, reduction in composite properties due to the low penetrability of the cement matrix i.e., low bonding, as the size of the bundle increased. For a composite with a smaller bundle diameter, the situation is opposite, the low reinforcement content reduces the composite performance; however, the improved cement penetrability and bonding increases the entire composite performance. Therefore, no significant difference is reported for the two composite systems with the different bundle size at relatively low strain values (Specimens 4 and 6 in Table 3.2). The efficiency factors of the different bundle systems support the above discussion. The efficiency factors can enable a comparison of the different systems while taking into account the volume content of reinforcement as well as fabric properties. Table 3.2 shows that the efficiency factor of Specimen 4 with the large size bundle (138) is much lower, about 0%, than that of Specimen 6 with the thinner diameter bundle (88 tex) for low strain values of 1% and 2%. At the higher strain value (at failure) no significant difference in bundle efficiency is observed ( ) HDPE - 4mm loop size Tensile Stress [MPa] tex 88 tex Fig. 3. 2: 0 0% 2% 4% 6% 8% Strain Stress-Strain responses of composites with HDPE having different tex values and 4 mm loop size

8 30 COHEN,PELED,PASDER,ROYE,GRIES: Effects of warp knitted fabrics made from multifilament The above discussion can also explain why there is only a slight difference between the tensile behavior of the composites with the two bundle sizes (Fig. 3.3) when the fabric is made from a smaller loop size of 2 mm (Specimens 3 and in Table 3.2). Also in these cases the bond strengths as well as the reinforcing efficiency are greater for the small sized bundle but here, the difference in tensile behavior between the low and high bundle diameter systems is much smaller than that observed with the 4 mm loop size fabrics (Fig. 3.2). Here, the stitches are denser and smaller and therefore strongly tightening the filaments in the bundle. Such tightening effects can prevent cement penetrability in between the filaments of the bundle, as schematically described in Fig With the large size bundle such a reduction in the cement penetrability can be a dominant factor leading to a significant reduction in the composite performance, even at high strain levels. However, with the small diameter bundle (88 tex) due to its small size, these tightening effects are not as dominant. Therefore,, for the 2 mm loop size system, the difference in tensile performance between the high and small size bundle fabric systems is not as significant as in the case with the more open bundle with 4 mm loop size. Tensile Stress [MPa] tex 88 tex HDPE - 2mm loop size 0 0% 2% 4% 6% 8% Strain Fig. 3. 3: Stress-Strain responses of composites with HDPE having different tex values and 2 mm loop size Fig. 3. 4: -cement -fibers/pores Illustration of the cement penetration into dense (right) and undense (left) yarn 3. Influence of loop size The influences of the loop size on the tensile behavior of HDPE knit fabric are shown in Fig. 3. for low and high size bundles (yarn tex). The comparison shows a significantly higher strength for the 4 mm loop size fabric than that with 2 mm loop size for bundles with 138 tex (Samples 3 and 4 in Table 3.2 and Fig. 3.b), mainly at large strains. This trend is not as significant when comparing the composite tensile properties of the 88 tex bundle, here, the small loop size system even shows some improvement in the tensile stresses. The efficiency factors as well as bond strengths are greater for the 4 mm loop size systems for both bundles

9 Tensile Stress [MPa] 1st International Conference Textile Reinforced Concrete (ICTRC) 31 systems, with 88 tex and 138 tex, respectively. In general, the improvement in the efficiency factor of the 4 mm loop size composite over the composite with 2 mm loop size is greater for the 138 tex system, above 0%, than that of 88 tex, which exhibits less than 0% improvement. The influences involved in the size of loops of the knit fabric are similar to those discussed in the previous chapter. The increase in loop size in a knit fabric can result in several opposite influences: (i) increasing the number of loops may provide stronger anchorage with the cement paste as the number of anchoring points increases; (ii) on the other hand, the stitches tighten the filaments in the bundle and prevent spaces from being opened to the outside, thus they reduce the matrix penetrability. As the number of stitches increases, their size decreases, further reducing the matrix penetrability and overall composite performance; and (iii) the tensile strength of the fabric itself increases as the number of loops per mm decreases and their size enhances, i.e., 4 mm is stronger than the 2 mm (Table 3.1). The optimal loop size is a combination of these different factors. Based on the above discussion, it can be concluded that a thick bundle (high tex number) with a large diameter is more sensitive to cement penetrability and therefore highly affected by the loop size. Increasing the size of the loop helps with cement penetrability and can improve composite performance. However, when a thinner bundle is used to produce the knit fabric, cement penetrability is much easier and therefore, the influence of the loop size is less significant, as observed in this work. At the small bundle yarn (88 tex) the improved performance of the composite with the 4 mm loop size takes place mainly at the stage where the loads are carried by the fabrics (Table 3.2) HDPE - 88 tex 2 mm 4 mm 0 0% 2% 4% 6% 8% Strain Fig. 3. : (a) Tensile Stress [MPa] HDPE tex 2mm 4mm 0 0% 2% 4% 6% 8% Strain Comparison of composite tensile behavior reinforced with HDPE fabrics made with different loop size and bundle size: (a) with 88 tex, and (b) with 138 tex (b) 4 SUMMARY AND CONCLUSIONS In warp knitted fabrics, warp and weft yarns are connected by stitches (loops). Decreasing the number of loops per unit length in a knit fabric improves the fabric's mechanical performance, as observed in this work. On the contrary, a larger number of loops per unit length provides more anchoring points and thus stronger anchorage with the cement paste. However, stitches tighten the filaments in the bundle and prevent spaces from being opened, thus they reduce the matrix penetrability. As their number increases, their size decreases, further reducing the

10 32 COHEN,PELED,PASDER,ROYE,GRIES: Effects of warp knitted fabrics made from multifilament matrix penetrability and overall composite performance. The optimal loop size is a combination of these different factors. Based on the present work, it can be concluded that a knit fabric produced with a large diameter bundle is more sensitive to cement penetrability and therefore highly affected by the knitted fabric loop size. In this case, the optimal loop size was found to be 4 mm and therefore provides the best composite performance. When a thinner bundle is used to produce the knit fabric, cement penetrability is much easier and therefore, the influence of the loop size is less significant. A complementary study is needed in order to get more accurate conclusions regarding influences of knit fabric structures including loop size on bonding, cement penetrability and composite performance, as well as development of a mathematical model. The influence of different fabric materials shows a high strengthening of TRC with Aramid and HDPE yarns, mainly at large strains. The results brought in this study indicate the massive potential uses of those TRC components, such as airport pavement, dams, construction in places endangered by earthquakes and many more. REFERENCES [Ban04] Banholzer, B. and Brameshuber, W. Tailring of AR-glass filament/cement based matrix bond analytical and experimental techniques, p , [Ben97] Bentur, A., Peled, A., Yankelevsky, D., "Enhanced bonding of low modulus polymer fibers-cement matrix by means of crimped geometry", J.Cement & Concrete Research, 27(7), p , [Del86] Delvasto S., Naaman A.E., and Throne J.L., "Effect of pressure after casting on high strength fiber reinforced mortar", Int. J. Cement Composites Lightweight Concrete, 8(3), p , [Jes03] Jesse, F, and Curbach, M. Strength of continuous AR-glass fiber reinforcement of cementitious composites. Fourth International Workshop on High Performance Fiber Reinforced Cement Composites (HPFRCC 4), RILEM Publications, p , [pel98a] Peled, A., Bentur, A., Yankelevsky D., "Effects of woven fabrics geometry on the bonding performance of cementitious composites: mechanical performance", J. Advanced Cement Based Materials, 7, p , [Pel99b] Peled, A., Bentur, A., Yankelevsky D., "Flexural performance of cementitious composites reinforced by woven fabrics", J. Materials in civil Engineering (ASCE), p , [Pel00c] Peled, A., Bentur, A., "Geometrical characteristics and efficiency of textile fabrics for reinforcing composites", J. Cement and Concrete Research, 30, p , [Pel00d] Peled A. and Mobasher B., "Pultruded Fabric-Cement Composites" ACI Materials Journal, 102(1), p. 1-23, [Pel03e] Peled A. and Shah S.P., "Processing Effects in Cementitious Composites: Extrusion and Casting", Journal of Materials and Civil Engineering, ASCE, p , March-April [Pel03f] Peled A. and Mobasher B., "Cement Based Pultruded Composites with Fabrics", Proceedings, 7th International Symposium on Brittle Matrix Composites (BMC7), Warsaw, Poland, p. 0-14, [Rei03] Reinhardt, HW, Kruger, M., and Grosse, U. Concrete prestressed with textile fabric. J. of Advanced Concrete Technology, 1(3); 231-9, [Roy04] Roye, T., Gries, and A., Peled, "Spacer fabrics for thin walled concrete elements", 6th RILEM Symposium on Fiber-Reinforced Concretes (FRC), BEFIB, Varenna, Italy, p , 2004.