Impact Behavior of 3D Fabric Reinforced Cementitious Composites A. Peled1, D. Zhu2, and B. Mobasher2 1 2

Structural Engineering Department, Ben Gurion University of The Negev Department of Civil and Environmental Engineering, Arizona State University

Abstract. The objective of this research was to study the behavior of 3D AR glass fabric cement-based composites under impact loading. It was found that 3D fabrics significantly improve the toughness and energy absorption of cementbased composites under impact loading, compared to short AR glass fibers reinforcement. The 3D fabric improves the toughness in as high as 200 folds compared to short fiber composites. The energy absorption was highly affected by the thickness of the element and the location of the 3D fabric faces. Greater toughness was obtained when the fabric faces were located in the direction of the hammer drop.

1 Introduction Cementitious materials may be subjected to dynamic loading in a variety of situations. Due to the inherent brittleness and low tensile strength of most cementbased elements, impact loadings can cause severe damage, resulting in extensive cracking. Since cement based composites are rate-dependent, their mechanical properties are highly dependent on the loading rate. Characterization of the impact response of the concrete is important for planning this activity. Textiles reinforced cement-based elements (TRC) have been intensively investigated in recent years [1]. Superior tensile strength, toughness, ductility and energy absorption were reported with TRC [2-3], properties which are important under dynamic loading. Several researchers recently showed the high potential of cement-based composites reinforced with fabrics under high loading rate [4-6]. Modern textile technology enables wide variety of fabric structures which allows great flexibility in fabric design. It is also possible to produce three-dimensional (3D) fabrics, providing reinforcement in the plane normal to the panel. The research to date focused mainly on the mechanical behavior of cement-based composites reinforced with two dimensional fabrics (2D). 3D fabrics having reinforcement in three orthogonal directions can limit failure by delamination, enhance shear strength and therefore expected to improve the mechanical properties of cement composites under dynamic and impact loads. Recently 3D spacer fabrics were developed for use in cement–based products. Several studies reported the G.J. Parra-Montesinos, H.W. Reinhardt, and A.E. Naaman (Eds.): HPFRCC 6, pp. 543–550. © RILEM 2012

544

A. Peled, D. Zhu, and B. Mobashher

efficiency of this fabric teechnology [7]. Limited studies dealt with the behavior oof cement-based compositess with 3D fabric as reinforcement [8-9], demonstratinng mainly the potential of using these types of reinforcements in the cements field. The objective of this reesearch was to study and compare the behavior of fabriccement composites rein nforced with 3D fabrics, under impact loading. Thhe influences of fly ash as well w as specimen position were studied. The results werre compared to cement comp posites reinforced with short AR glass fibers.

2 Experimental 2.1 Preparation of Specimens S 3D warp knitted fabric structures s were used in this work, in which two sets oof independent 2D knitted fabrics f were connected together with a third set of yarnns along the thickness, Z diirection, of the fabric, referred as spacer yarns. The 3D m fabric is presented in Fig. 1a. The yarns in the X and Y directions were made from alkali resistance (AR) glass bundles of 1200 tex with tensile strength of 1325 MP Pa and 67 GPa modules off elasticity, these yarns were the reinforcing yarns annd connected together by stittches (loops) made of fine polyester. The volume contennt, Vf, of AR glass reinforceement yarns was of ~1% at the composite plain at botth directions. The spacer yarrns in the Z direction were used mainly for stabilizatioon of the 3D fabric. The 3D D fabrics were developed and produced by ITA, RWT TH Aachen.

(a)

(b)

(c)

10 mm Fig. 1. (a) 3D fabric struccture and composite testing arrangements: (b) horizontal, ((c) vertical

Two 3D fabric compo osites were prepared one with cement paste (water annd cement only, 3DGPC) off 0.4 water/cement ratio and another with replacement oof 20% by volume of cemeent by fly ash (3DGFA). The 3D fabric specimens werre prepared by placing the faabric in a mold and then casting the cement matrix on toop of the fabric, until compllete covering of the fabric. Vacuum was applied durinng casting to allow good penetrability of the cement matrix into the opening of thhe 3D fabric. 24 hours after casting % c the specimens were demolded and cured in 100% relative humidity for 7 daays and then at room environment until testing in impacct,

Impact Behavior of 3D Fabric Reinforced Cementitious Composites

545

28 days after casting. Before testing, each composite plate was cut to slices providing specimens with a 200 x 50 x 25 mm of length, width and thickness, respectively. For comparison, composites with short AR glass fibers of 12 mm length were prepared (GF12) with similar dimensions and procedure as the 3D fabric composites. The fiber content was 0.3% by weight.

2.2 Impact Test Procedure An impact test set-up based on a free-fall drop of an instrumented hammer on a three point bending specimen was used with a span of 178 mm and load cell with a range of 90 kN. The hammer weight was 134 N with a drop height of 152 mm. A linear variable differential transformer (LVDT) with a range of +10 mm was connected to the specimen by means of a lever arm. A high speed digital camera (Phantom v.7) was used to capture pictures of the samples during the impact tests. The damage caused in the samples was then compared by visual examination. The reinforcing direction of the fabric was along the stitches for all systems. The toughness, maximum stress and rigidity were calculated. The toughness was calculated by the energy dissipated in the specimen using the area under the load-deflection curves. The maximum impact stress was calculated using linear elastic small displacement bending equation and the rigidity was calculated by the slope of load-deflection curve in the linear elastic region. Typical stresses vs. time and stresses vs. deflection curves were chosen to compare between the different tested systems. For more details on the impact testing see [5].

3 Results and Discussion 3.1 Test Direction Two loading arrangements were carried out: parallel to fabric layers, horizontal arrangements (Fig. 1b) and perpendicular to the fabric layers, vertical arrangement (Fig. 1c). In the horizontal arrangement the two faces of the 3D fabric are located at the top and bottom of the specimen (as it appears in Fig. 1a) relative to the drop direction of the hammer, where the spacer yarns are passing through the thickness from top to bottom. In the vertical arrangement the two faces of the 3D fabric are located at the sides of the specimen relative to the drop of the hammer, and the spacer yarns are passing through the width of the composite between the two sides of the composite. The average results are presented in Table 1.

546

A. Peled, D. Zhu, and B. Mobasher

The impact stress vs. deflection curves of the horizontal and vertical arrangement specimens are presented in Figs. 2a and 2b. A uniform impact behavior is observed, as for the horizontal arrangement this uniformity is even clearer. This indicates the uniformity of the composites and the ability of using this test to study the impact behavior of the TRC systems. It is also seen that the deflection of the TRC systems is very high exhibiting values of up to about 16 mm. A rebound behavior of the specimens tested at vertical arrangement is clearly seen, i.e., after reaching the maximum deflection the specimen was shifted back to some extent and no failure has occurred.



8

Horizontal

3DGPC

Stress, MPa

6 4 2 0

(a)

rebound rebound 0

4 8 12 D eflection , mm

16

(b)

8 Horizontal Vertical

Stress, MPa

6 4 2

rebound 0

(c)

(d)

0

0.01

0.02 Time, sec

0.03

Fig. 2. Impact behavior of 3D fabric composites tested at (a) horizontal and (b) vertical arrangements, comparison of testing arrangements, (c) stress vs. deflection and (d) stress vs. time

Comparison of the typical curves of each system is presented in Figs. 2c and 2d, showing the impact stresses vs. deflection and vs. time. The rebound mechanism is obvious for the vertically tested system. On the other hand not such rebound, i.e. shifting back behavior, is seen for the composites tested at horizontal arrangement. Due to this rebound behavior the entire duration of the test is greater, about twice as much for the composite tested vertically as compared to the composite tested horizontally. The toughness of the vertical system is also much greater than that of the horizontal system (Table 1), providing much better energy absorption of the vertically tested composite. However, the horizontal system exhibits greater maximum stress as compared to the vertical system. But the

Impact Behavior of 3D Fabric Reinforced Cementitious Composites

547

improved strength by the horizontal arrangement is much smaller, of about 15%, than the improved toughness by the vertical system, of about 50%. So when energy absorption is considered the thickness of the element and location of the fabric faces relative to the impactor, are important factors. Table 1. Impact properties of the composites Compo- Sample site dir.

Rigidity, N/mm

3DGFA Horiz.

27223±4273

744±73

3DGPC Horiz.

11879±1075

876±49

GF12

Max. Max. Stress, Force, N MPa

Max. Deflection, mm

Deflection at Toughness, Max. Force, N.mm mm

5.15±0.30

5.98±0.53

0.143±0.179

6.33±0.93

12.05±2.21

0.639±0.338 5460±1110

Vert.

11014±245 1281±124 5.38±0.49

11.90±2.21

0.824±0.288

8100±961

Horiz.

29685±5147

0.077±0.005

0.037±0.005

31.3±3.6

730±67

5.68±0.55

2255±379

The crack pattern and development of these two systems is compared in Fig. 3. When tested horizontally, at duration of 0.75 ms the developed crack is getting almost through the entire thickness of the composite (Fig. 3a). At latter stages of the impact test the crack is widening, and the opening of the crack is relatively large at duration of 13.25 ms (Fig. 3b). The reinforcing yarns of the 3D fabric are pulling out and holding the specimen from a complete failure as seen in Fig. 4a, a major damage of the whole composite is obvious. Broken bundles and filaments at the tensile zone of the composite can easily be seen in Fig. 4b. These last two images were taken after the impact test was ended. However for the vertically tested system the crack development is different, at early test duration of 0.75 ms very fine crack is observed which developed only up to the middle of the composite thickness (Fig. 3c). For this composite at late stages the crack developed through most of the composite thickness but the reinforcing yarns of the fabric are bridging the crack through the specimen thickness and the composite did not fail (Fig. 3d).

Crack

Vertical

Horizontal (a)

t = 15.75 ms

t = 0.75 ms

t = 0.75 ms Crack

(b)

(c)

(d)

Fig. 3. Comparison of 3D fabric composites tested at two arrangements at different durations

When observing the vertically tested composite after the impact testing (Fig. 4c), the crack is much finer, as part was closed due to the rebound mechanism, discussed above. No significant damage of the fabric and yarn brakeage was obtained with this composite at the end of testing, with no major failure of the composite. This composite can still hold loads and remain generally in its original shape after ending the test with the specific test parameters discussed above.

548

A. Peled, D. Zhu, and B. Mobashher

(b)

(a)

(c)

Fig. 4. Comparison of 3D faabric composites at two test arrangements at the end of testinng: (a) horizontal, (b) horizontall after impact test (c) vertical after impact test

3.2 3D Fabrics vs. Short S Yarns Comparison of the short ffiber composites with the 3D fabric composites is show wn in Fig. 5. This figure com mpares typical curves of the 3D fabric composite with thhe two matrices: with fly ash (3DGFA) and without fly ash (3DGPC) and the shoort fibers (GF12), presenting me g stress vs. deflection at log scales and stress vs. tim linear scale. The test arrrangement in all is horizontal. The significant britttle behavior of short fiber co omposites is obvious as compared to much more ductiile behavior of the 3D fabric composites. Also the duration of the test is much smalleer for the fiber composites as compared with the 3D fabric composites. The shoort fiber composite duration last l less than 0.1 ms with a maximum deflection of abouut 0.08 mm, as for the 3D fabric f composites the test duration is about 0.01 ms witth deflection of about 10 mm m. 7

7

6

6

5

5

4

4

3

3

2

2

1

1

0 0.001

(a)

0.01

0.1

1

Deflectionn, mm

10

0

100

(b)

0

0.003

0.006

0.009

0.012

Time, sec

Fig. 5. Comparison of comp posites impact behavior with short glass fibers and 3D fabrics with plain cement and fly assh matrices: (a) stress vs. deflection logarithm scale, (b) streess vs. time

Such small deflection leads l to very low toughness of the short fiber compositees of 31 N mm as compareed to much greater toughness of 4560 N mm of the 3D fabric composite with th he plain matrix (Table 1). Crack pattern is presented iin Fig. 6. Wide crack throug gh the entire thickness of the short fiber composite waas

Impact Behavior of 3D Fabric Reinforced Cementitious Composites

5449

developed at ~9 ms of teest duration (Fig. 6b). No fibers bridging the cracks arre seen at this stage of loadiing. A complete failure of the composite has occurred at the end of the impact test of the short fiber composite (Fig. 6c) correlating with thhe brittle behavior obtained d with these composites. Note that the 3D fabric alsso slightly improves the maaximum impact stress, of about 10%, compared to thhe short fiber composite, bu ut this improvement is significantly smaller than that oof the toughness of almost 200 2 folds. Based on that, the benefit of the 3D fabric aas reinforcement for cement-based composites is improving the energy absorption oof the composite. t = 0.75 ms

(a)

t = 9.25 ms

(b)

(c)

Fig. 6. Comparison of imag ges of short fiber composite at different duration of testing: ((a) 0.75 ms, (b) 9.25 ms, and (c)) end of testing.

Addition of fly ash results r in reduction in composite properties. The moost significant reduction was obtained for the toughness; the fly ash compositees exhibit almost half of thee toughness as that of the plain paste composites, witth values of 2255 N.mm and d 5460 N.mm for the fly ash and no fly ash compositees, respectively. Similar tren nd was also observed for the deflection. Further thhe presence of the fly ash sliightly reduces the maximum stress from 6.33 MPa of thhe plain paste matrix to 5.1 13 MPa of the fly ash matrix. The fly ash might causse reduction in the bonding between the 3D fabric and the cement matrix, leading at least to some extent to th he reduction in composite performance. However furtheer work is required to better understand this behavior.

4 Summary and Co onclusions In this work the impact prroperties of 3D AR glass fabric cement-based compositees were studied and compared to short AR glass fiber composites. It was found thhat 3D fabrics significantly im mprove the toughness and energy absorption of cemenntbased composites to as hig gh as 200 folds compared to short fiber composites. The arrangement of th he fabric within the composite and composite dimensioon relative to the dropped haammer found to have significant influence on the impaact behavior. Improved tough hness in about 50% was obtained when the fabric waas placed in the direction of o the dropped hammer as compared to the horizonttal arrangement. When energ gy absorption is considered the thickness of the elemennt and location of the 3D fab bric faces relative to the impactor, are important factors.

550

A. Peled, D. Zhu, and B. Mobasher

Based on this work in can be concluded that 3D fabric can be beneficial as reinforcement for cement-based composites. Further work is required in order to better understand and design these composites. In this work the yarns in the Z direction were spacer yarns with no real reinforcing abilities, for getting better performance stronger yarns are suggested along the Z direction. Acknowledgements. The authors would like to thank ITA at RWTH Aachen University for their cooperation for providing the 3D fabrics used in this study.

References [1] Brameshuber, W. (ed.): Textile reinforced concrete. State of the art report. RILEM TC 201-TRC, RILEM (2006) [2] Peled, A., Cohen, Z., Pasder, Y., Roye, A., Gries, T.: Influence of textile characteristics on the tensile properties of warp knitted cement based composites. Cement and Concrete Composites J. 32, 174–183 (2008) [3] Peled, A., Mobasher, B.: Tensile Behavior of Fabric Cement-Based Composites: Pultruded and Cast. ASCE, J. of Materials in Civil Engineering 19(4), 340–348 (2007) [4] Zhu, D., Peled, A., Mobasher, B.: Dynamic tensile testing of fabric-cement composites. Construction and Building Materials J. 25(1), 385–395 (2011) [5] Zhu, D., Gencoglu, M., Mobasher, B.: Low velocity flexural impact behavior of AR glass fabric reinforced cement composites. Cement and Concrete Composites J. 31, 379–387 (2009) [6] Haim, E., Peled, A.: Impact behavior of fabric-cement hybrid composites. ACI Materials J. 108(02) (March - April 2011) [7] Hanisch, V., Kolkmann, A., Roye, A., Gries, T.: Influence of machine settings on mechanical performance of yarns and textile structures. In: Hegger, J., Brameshuber, W., Will, N. (eds.) Proceedings of the 1st International RILEM Symposium (Textile Reinforced Concrete ICTRC) RILEM TC201-TRC, pp. 13–22 (2006) [8] Roye, A., Gries, T., Peled, A.: Spacer fabric for thin walled concrete elements. In: Di Prisco, M., Felicetti, R., Plizzari, G.A. (eds.) Fiber Reinforced Concrete – BEFIB, PRO 39, RILEM, pp. 1505–1514 (2004) [9] Naaman, N.: Textile reinforced cement composites: competitive status and research directions. In: Brameshuber, W. (ed.) International RILEM Conference on Materials Science (MatSci) I, pp. 3–22 (2010)