performance of fiber reinforced materials: historic ... · ideally high performance ﬁber...

Performance of Fiber Reinforced Materials:Historic Perspective and Glance in the Future

Surendra Shah(&) and Yuan Gao

Civil and Environmental Engineering, Northwestern University,Evanston, IL 60208, USA

[email protected]

Abstract. Ideally high performance fiber reinforced concrete should haveenhanced modulus of elasticity, higher bend over point, sustained strain hard-ening response characterized by sequential multiple cracking with controlledcrack widths. Conventional fiber reinforced concrete primarily alters the postpeak response and can constrain macro cracks. Strain hardening response can beobtained by controlling processing (extrusion, pultrusion, spraying chopped fiberand cement slurry), volume, type and geometry of fibers (e.g. textile) and rhe-ology of matrix (self-consolidating concrete (SCC), ultra-high performanceconcrete (UHPC)). Use of nanofibers such as carbon nanotubes (CNT) and car-bon nanofibers (CNF) fundamentally alters the nano structure of calcium silicatehydrate (C-S-H) gel. Reinforcing concrete with CNT can enhance modulus ofelasticity, enhance toughness, reduce autogenous shrinkage and delay corrosionof steel bar reinforcement. Modeling and characterizing interfaces at multiscale iscritical in development of high performance fiber reinforced concrete.

Keywords: Textile � Carbon nanotube � Multiply cracking � Crack width �Autogenous shrinkage � Interface � Debonding

1 Introduction

The current development of fiber reinforced concrete can be traced to two sources.In U.S., the use of randomly distribute steel fibers was launched by Romualdi andBatson (Romualdi and Batson 1963). They argued that one can substantially increasetensile strength of concrete by closely spaced steel fibers. They mentioned that shortsteel fibers can be closely spaced, and based on Griffith Fracture Criteria, the smallerthe spacing of reinforcement, the shorter the crack length, and the higher the fracturestrength. This assumes that the cracks are of the same size as the spacing of fibers.However, the fracture process of concrete starts with microcracks, and by the time thecracks interact with fibers, microcracks have coalesced into macro crack. Thus it islikely that fibers which are in millimeter range do not influence the tensile strength.This was shown by Shah and Rangan (Shah and Rangan 1971). They showed thatmacro fibers do not increase the tensile strength but do substantially enhance toughness(Figs. 1). This is now the widely accepted response of conventional fiber reinforcedconcrete where fibers are in millimeter range and the volume of fibers is in the range of1% (Balaguru and Shah 1992).

© RILEM 2018V. Mechtcherine et al. (eds.), Strain-Hardening Cement-Based Composites,RILEM Bookseries 15, DOI 10.1007/978-94-024-1194-2_1

2 Conventional Fibers

The major influence of fiber reinforced concrete is in enhancing toughness, in con-trolling crack width and altering the post-peak response of stress–strain curve in tensionor load-deflection response in flexure. How to measure fracture toughness? For aclassical brittle material, fracture toughness can be evaluated using liner elastic fracturemechanics. However, concrete and fiber reinforced concrete are quasi-brittle material(Shah et al. 1995). The peak tensile stress is proceeded by a stable growth of micro-cracks (often called fracture process zone). This non-linear behavior means that fracturetoughness evaluated using linear elastic fracture mechanics (LEFM) or by measuringthe area under the load-deformation curve has strong size and geometry dependence.Several possibilities exist to account for the existence of fracture process zone. Theseinclude: Jenq and Shah’s two parameter model (Jenq and Shah 1985), Hillerborgfictitious crack model (Hillerborg 1978) and Bazant’s size effect law (Bazant andPlanas 1997). Similarly, several empirical test methods have been proposed. The mostrational one appears to be the measurement of load-CMOD curve using notched beam(BSI 2005).

In addition to steel fibers, other types of fibers have been successfully used forcontrolling cracks. One major application is in slabs or pavements to control shrinkagecracking. The restrained ring test is frequently used to compare performance of variousfibers (Grzybowski and Shah 1990).

Fig. 1. Effect of spacing of reinforcement on cracking strength of concrete (Shah and Rangan1971).

4 S. Shah and Y. Gao

The second source of the current development of fiber reinforced concrete camefrom the need to replace asbestos fibers in so-called cement board. This led to thecurrent use of glass fiber reinforced cement (GFRC). These GFRC panels are producedby spraying glass fiber and cement slurry on a desired form. The amount of choppedglass fibers is in the range of 4–12%. The response of a typical GFRC in tension isshown in Fig. 3. It can be seen that after the elastic response first matrix crackingoccurs (macro cracks). This point is often called bend over point (BOP). After BOP,strain hardening type of response is observed. Unless precautions are taken in for-mulation of glass fibers added cement matrix, unfavorable aging can occur and thedesirable strain hardening response may disappear.

Asbestos fiber reinforced sheets are made with so-called Hatschek process. In thisprocess, slurry of cement and fibers are deposited layer by layer. Asbestos fibers arenow replaced with cellulose fibers, polypropylene (PE) fibers or polyvinyl alcohol(PVA) fibers. The volume of fibers in this additive manufacturing process is consid-erably more than those used in conventional cast-in-place construction (Kuder andShah 2003). Cement boards made with Hatschek process are widely used in many partsof the world.

With cast-in-place construction, it is often difficult to add more than about 1% byvolume of fibers. Addition of fibers make concrete less workable. Depending upon thetype and geometry of fibers, and the rheology of matrix, the higher the amount of fibers,the more difficult it is to uniformly disperse them. One alternate to cast in place con-structions is extrusion. A stress–strain curve of fiber reinforced extruded tensile panel isshown in Fig. 4. The matrix was cement paste and fibers were PVA fibers (14 micron indiameter, 2 mm long, and 2% by volume). It can be seen that the addition of fibers hasenhanced the bend over point and substantial strain hardening response is obtained.

Fig. 2. Average crack width versus fiber volume (Grzybowski and Shah 1990).

Performance of Fiber Reinforced Materials 5

The ability to extrude depends upon the ability of matrix to flow through a given dieshape and maintain dimensional control immediately after being extruded. The shapestability is controlled by controlling thixotropy. Addition of nano clay has been shown tofacilitate extrusion (Kuder and Shah 2007). It should be mentioned that the knowledgegained from Hatschek process and extrusion should be helpful in 3D printing elements.

Fig. 3. Typical load-elongation curve for glass fiber in tension (Shah et al. 1987).

Fig. 4. Tensile stress–strain curve of extruded cement composite (Shao and Shah 1997).


In addition to extrusion, Mobasher and his colleagues have also successfully madecomposites with pultrusion (Peled and Mobasher 2005).

In extruded composites, no aggregates are used. This made possible to use largeamount of fibers. Ultra-High Performance Concrete (UHPC) is made with only sandsize particles. This makes it possible to use relatively large volume of fibers and obtainstrain hardening type of response. Uniform dispersion of fibers is facilitated in UHPCby making matrix more fluid. This is similar to self-consolidating concrete (SCC).With SCC, a higher amount of fibers can be and have been used. It has been also shownsince SCC does not involve vibration, the dispersion of fiber is more uniform (Ozyurtet al. 2007).

Currently there is a considerable interest in textile reinforced concrete for both newconstruction and repair of existing structures such as brick masonry. Perhaps the first

Fig. 5. Load-CMOD curve for cement mortar with CNTs or CNFs.

Fig. 6. Restrained ring test of cement mortar with or without CNF.


use of textile reinforcement is ferrocement when mortar is reinforced with steel wiremeshes. Pier Luigi Nervi extensively used ferrocement for shell structures. He showedthat even 10 mm thick ferrocement boats can be water tight. Naaman and Shah (1971)showed that crack spacing and crack width are related to the specific surface of wiremesh reinforcement.

In concrete structures, concrete is reinforced with steel bars or prestressing strands.The rebars or strands do not alter tensile strength (BOP) of concrete. Only after themajor cracks are formed, reinforcing bars constrain these cracks so that the ultimateload is much higher than the load at cracking. High performance fiber reinforced

Fig. 7. Young’s modulus of cement composites with or without CNTs.

14 16 18 20 22 24 26 28 300

5

10

15

20

25

30

35

40

45

Cou

nt

Modulus (GPa)

Without CNFs With CNFs

Fig. 8. Modulus of interface between cement paste and sand from AFM test.


concrete should enhance the bend over point (that is increase the tensile strength of thematrix). This is possible since fibers can interact with microcracks and delay formationof macrocracks. Microcracks are initiated from flaws in cement paste matrix. Can weconstrain these flaws which are in nanoscale? These considerations have led to thedevelopment of concrete reinforced with carbon nanotube, graphene sheets and gra-phene oxide and carbon nanofibers.

3 CNTs and CNFs

There are two challenges with the addition of carbon nanotubes (CNTs) and carbonnanofibers (CNFs). One is dispersion and the other is cost. Researchers at Northwesternhave solved these problems by uniformly dispersing CNT and CNF and efficiently

Fig. 9. Interface debonding and sliding (Vf = 1.3%) (Shao et al. 1993).


using only a very small amount (less than 0.1%). The method of uniform dispersioninvolves the use of surfactant and ultrasonication (Konsta-Gdoutos 2010).

Addition of well dispersed CNT or CNF have some very interesting results. Forexample, the load-CMOD curves of reinforced and control matrix are shown in Fig. 5.The addition of up to 0.1% fibers has increased the modulus of elasticity, flexuralstrength and post peak response.

In addition to enhance mechanical properties, the addition of CNT was shown toreduce autogenous shrinkage. The results of restrained ring test show that CNT rein-forced matrix not only delays onset of cracking but reduces crack width (Fig. 6).

The enhanced modulus with CNT addition has caught the interest of the structuraldesigners of tall buildings. The design of tall buildings is often governed by deflectionsand deformations. So it is desirable to have concrete with very high modulus. Theincrease in modulus due to 0.1% addition of CNT for cement paste, mortar and con-crete is shown in Fig. 7. It is surprising and unexpected that the increase in modulus isproportionally more in concrete than in paste. This cannot be modeled with a con-ventional composite material approach. It is likely that addition of CNT has enhancedthe modulus of interface at nanoscale. This is shown in Fig. 8.

Addition of silica fume also densifies interface. However, unlike silica fume theaddition of CNT does not significantly increase compressive strength nor does itincrease brittleness.

4 Conclusion

Ideally, high performance fiber reinforced concrete can have the following attributes:(1) Enhanced modulus of elasticity, (2) higher bend over point, (3) strain hardening

response which is characterized by sequential multiple cracking and crack widths inmicron range.

Fibers at different scale have different roles to play in assuring the desired response.Fibers at nanoscale (CNT and CNF) can constrain nano scale voids and flaws. Theycan alter the nano structure of calcium silicate hydrate, improve pore structure andenhance modulus as well as transport properties. In addition, CNT and CNF have beenshown to control autogenous shrinkage cracking and increase corrosion resistance ofsteel reinforcement (Konsta-Gdoutos et al. 2017).

Fibers at micro scale can constrain microcracks and as a result enhance bend overpoint.

Fiber, matrix and interface parameters play a critical role in assuring strain hard-ening response. The importance of interfacial bond can be seen in Fig. 9. It can be seenthat even at the high value of strain in the strain hardening part, there is only partialdebonding, thus controlling crack width. Many attempts have been made at modelingthese above stages of response (Mobasher et al. 1991; Li et al. 1992; Ouyang et al.1997). Professor Mechtcherine and his group have more recently studied these aspects(Boshoff et al. 2016).


References

Balaguru, P.N., Shah, S.P.: Fiber-Reinforced Cement Composites. McGraw-Hill, New York(1992)

Bazant, Z.P., Planas, J.: Fracture and Size Effect in Concrete and Other Quasibrittle Materials.CRC Press, Boca Raton (1997)

BSI (British Standards Institution). Test Method for Metallic Fiber in Concrete. Standard BS EN14651. BSI, London (2005)

Boshoff, W.P., Altmann, F., Adendorff, C.J., Mechtcherine, V.: A new approach for modellingthe ingress of deleterious materials in cracked strain hardening cement-based composites.Mater. Struct. 49(6), 2285–2295 (2016)

Grzybowski, M., Shah, S.P.: Shrinkage cracking of fiber reinforced concrete. Mater. J. 87(2),138–148 (1990)

Hillerborg, A.: A Model for Fracture Analysis. Division of Building Materials, Lund Institute ofTechnology, Lund (1978)

Jenq, Y., Shah, S.P.: Two parameter fracture model for concrete. J. Eng. Mech. 111(10), 1227–1241 (1985)

Konsta-Gdoutos, M.S., Metaxa, Z.S., Shah, S.P.: Highly dispersed carbon nanotube reinforcedcement based materials. Cem. Concr. Res. 40(7), 1052–1059 (2010)

Konsta-Gdoutos, M.S., Batis, G., Danoglidis, P.A., Zacharopoulou, A.K., Zacharopoulou, E.K.,Falara, M.G., Shah, S.P.: Effect of CNT and CNF loading and count on the corrosionresistance, conductivity and mechanical properties of nanomodified OPC mortars. Constr.Build. Mater. 147, 48–57 (2017)

Kuder, K.G., Shah, S.P.: Effects of pressure on resistance to freezing and thawing offiber-reinforced cement board. Mater. J. 100(6), 463–468 (2003)

Kuder, K.G., Shah, S.P.: Rheology of extruded cement-based materials. Mater. J. 104(3), 283–290 (2007)

Li, S., Li, Z., Mura, T., Shah, S.P.: Multiple fracture of fiber-reinforced brittle matrix compositesbased on micromechanics. Eng. Fract. Mech. 43(4), 561–579 (1992)

Mobasher, B., Ouyang, C., Shah, S.P.: Modeling of fiber toughening in cementitious materialsusing an R-curve approach. Int. J. Fract. 50(3), 199–219 (1991)

Naaman, A.E., Shah, S.P.: Tensile tests of ferrocement. J. Proc. 68(9), 693–698 (1971)Ouyang, C., Wollrab, E., Kulkarni, S.M., Shah, S.P.: Prediction of cracking response of

reinforced concrete tensile members. J. Struct. Eng. 123(1), 70–77 (1997)Ozyurt, N., Mason, T.O., Shah, S.P.: Correlation of fiber dispersion, rheology and mechanical

performance of FRCs. Cem. Concr. Compos. 29(2), 70–79 (2007)Peled, A., Mobasher, B.: Pultruded fabric-cement composites. Mater. J. 102(1), 15–23 (2005)Romualdi, J.P., Batson, J.B.: Behavior of reinforced concrete beams with closely spaced

reinforcement. J. Proc. 60(6), 775–790 (1963)Shah, S.P., Rangan, B.V.: Fiber reinforced concrete properties. J. Proc. 68(2), 126–137 (1971)Shah, S.P., Daniel, J.I., Ludirdja, D.: Toughness of glass fiber reinforced concrete panels

subjected to accelerated aging. PCI J. 32(5), 82–99 (1987)Shah, S.P., Swartz, S.E., Ouyang, C.: Fracture Mechanics of Concrete: Applications of Fracture

Mechanics to Concrete, Rock and Other Quasi-Brittle Materials. Wiley, Hoboken (1995)Shao, Y., Li, Z., Shah, S.P.: Matrix cracking and interface debonding in fiber-reinforced

cement-matrix composites. Adv. Cem. Based Mater. 1(2), 55–66 (1993)Shao, Y.X., Shah, S.P.: Mechanical properties of PVA fiber reinforced cement composites

fabricated by extrusion processing. Mater. J. 94(6), 555–564 (1997)


performance of fiber reinforced materials: historic ... · ideally high performance ﬁber...

Documents