Fracture behavior of cement mortar reinforced by hybrid composite fiber consisting of CaCO3 whiskers and PVA-steel hybrid fibers

We added CaCO3 whiskers into polyvinyl alcohol (PVA)-steel hybrid fiber system to obtain multiscale hybrid fiber reinforced cement mortar (MHFRC). Three-point bending (3-p-b) tests were carried out on 64 notched beams to investigate fracture behavior of MHFRC. Double K Fracture Criterion (DKFC) and Work Fracture Method (WFM) were employed to obtain fracture parameters. Influence of the volume fraction of CaCO3 whiskers, content of PVA-steel hybrid fiber and water/ cement ratio (w/c) on fracture parameters are discussed. Addition of composite fibers consisting of CaCO3 whiskers and PVA-steel hybrid fibers could improve fracture behavior of MHFRC. As the content of CaCO3 whiskers increases, fracture parameters first increase and then decrease. Content of PVA-steel fibers also affect fracture behavior of the matrix. Thus, an optimum ratio between CaCO3 whiskers and PVA and steel fibers contents exist that provide the best fracture performance of cement matrix: 1 vol % of CaCO3 whiskers, 0.5 vol % of PVA fibers and 1.5 vol % of steel fibers (S15P05W10). Influence of the w/c value is also discussed. Fracture toughness (KIC) and fracture energy (GF) of S15P05W10 group decreased as w/c values increased. The synergy of fibers in S15P05W10 was evaluated quantitatively, and the results indicated positive synergy effect on unstable fracture toughness (Kun IC ) and fracture energy (GF) in cement matrix with higher w/c values.

Accepted Manuscript

Fracture Behavior of Cement Mortar Reinforced by Hybrid Composite Fiber

Consisting of CaCO3 Whiskers and PVA-Steel Hybrid Fibers

Mingli Cao, Chaopeng Xie, Junfeng Guan

Please cite this article as: Cao, M., Xie, C., Guan, J., Fracture Behavior of Cement Mortar Reinforced by HybridComposite Fiber Consisting of CaCO3 Whiskers and PVA-Steel Hybrid Fibers, Composites: Part A (2019), doi:

https://doi.org/10.1016/j.compositesa.2019.03.002

This is a PDF file of an unedited manuscript that has been accepted for publication As a service to our customers

we are providing this early version of the manuscript The manuscript will undergo copyediting, typesetting, andreview of the resulting proof before it is published in its final form Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain

1

fiber and water/ cement ratio (w/c) on fracture parameters are discussed Addition of composite

fibers consisting of CaCO3 whiskers and PVA-steel hybrid fibers could improve fracture behavior

of MHFRC As the content of CaCO3 whiskers increases, fracture parameters first increase and then decrease Content of PVA-steel fibers also affect fracture behavior of the matrix Thus, an optimum ratio between CaCO3 whiskers and PVA and steel fibers contents exist that provide the best fracture performance of cement matrix: 1 vol % of CaCO3 whiskers, 0.5 vol % of PVA fibers

and 1.5 vol % of steel fibers (S15P05W10) Influence of the w/c value is also discussed Fracture

toughness (KIC) and fracture energy (GF) of S15P05W10 group decreased as w/c values increased

The synergy of fibers in S15P05W10 was evaluated quantitatively, and the results indicated positive synergy effect on unstable fracture toughness (Kun

IC) and fracture energy (GF) in cement

matrix with higher w/c values Negative synergy was observed for the initial fracture toughness (K

ini

IC) Comprehensive reinforcing index (RI v) was introduced as the characteristic parameters of the

hybrid fibers Fracture parameters increased first and then decreased as RI v increased Furthermore, the microscope morphologies of CaCO3 whiskers, PVA and steel fibers in the cement matrix were shown These results helped to establish microscopic reinforcing mechanism of the hybrid fibers

in the cement matrix Based on the experiment results, empirical formulas, which taking into

account fiber factor RI v and matrix factor w/c, were proposed to calculate fracture parameters of

MHFRC

2

water/ cement ratio

HIGHLIGHTS

 3-p-b tests for multiscale hybrid-fiber reinforced cement mortar were performed

 S15P05W10 group showed higher fracture toughness and fracture energy

 Synergy of fibers in S15P05W10 group with different waters/cement ratio were evaluated

 Microscopic reinforcing mechanism of hybrid fibers was clarified

 Theoretical models were proposed to calculate fracture parameters

1 Introduction

Fibers can effectively improve mechanical properties of cement-based materials, such as low tensile strength, brittle failure, weak cracking resistance and poor energy absorption capability [1-3] Thus, fibers are widely used as reinforcing materials, particularly in cement-based materials [1,4] Traditionally, metallic, polymeric and natural fibers, especially steel fibers are widely implemented in cement-based materials to “arrest” the cracks [5-6] The advantages of steel fibers include restricting or delaying small cracks from developing into macro-cracks, fiber bridge the cracks in post-cracking stage, and enhancing ductility and energy absorption capability of the matrix [2,4] Recently, to improve properties of cement-based materials and to save cost, the concept of hybrid fiber was proposed, which involves, mixing fibers of various types and with different sizes into a hybrid fiber composite [7] Comparing with single fiber reinforced cement-based materials, hybrid fibers inherit advantages of the individual fibers making hybrid-fiber reinforced cement-based materials with superior performance [7] Hossain [8] used polyvinyl alcohol (PVA) and steel fibers to reinforce self-consolidating concrete (FRSCC), and studied its strength and fracture energy The result showed that the compressive/flexural/splitting tensile strength and fracture energy of FRSCC improved, and fracture energy gain was more significant than strength gain Almusallam [9] performed mode-I fracture tests on steel-Kevlar-polypropylene hybrid-fiber reinforced concrete (HFRC), and studied how these fibers affected fracture energy of HFRC The results indicated that replacing steel fiber with Kevlar or polypropylene fiber does not affect the fracture energy However, higher steel fiber content, increases the fracture energy Lawler [10] reported that microfibers could increase the mechanical properties of hybrid-fiber reinforced mortar in pre-peak, and that macro-fibers play a bridging role

of macro-cracks during the post-cracking

3

However, all these hybrid fibers were macroscopic in scale Geometrical sizes of these fibers

are not compatible with the scales of cement hydration and cement pastes [11] Carbon nanotubes

and nanofibers were also reported for reinforcement of cement-based materials to achieve

toughening and reinforcing at the microscale [12-16] However, their complex dispersion

processes and high costs restrict their wider applications

CaCO3 whiskers are new type of micro-fibers able to effectively improve mechanical

properties of cement-based materials [17-19] Cao et al [20-22] employed CaCO3 whiskers, PVA

and steel fibers to form a new multiscale hybrid fiber system with multilevel and multiscale

characteristics equal or comparable to cement-based materials A series of test demonstrated that

cementitious composites reinforced with multiscale hybrid fibers could arrest cracks at multilevel

and improve the flexural strength, energy absorption capacity and plastic shrinkage performance

of the matrix Table 1 shown the types and size of different fibers used in the literatures

Table 1 The types and size of different fibers used in the literatures

Hossain et al [8]

FibraFlex (FF) metallic fiber Width 1.6mm, Thickness 0.029mm, Length 20mm

Width 1.0mm, Thickness 0.024mm, Length 5mm Macro- reinforcers

Almusallam et al [9]

Polypropylene fiber Width 1.0mm, Thickness 0.6mm, Length 50mm Macro- reinforcers

Konsta et al.[12][14] Multi-walled carbon nanotubes

(MWCNTs)

Diameter 20-40nm, Length 10-30μm Diameter 20-40nm, Length 10-100μm Nano-reinforcers

Fracture behavior is a fundamental mechanical property of hardened cement-based materials

that play an important role in design and safety evaluation of structures [23] Linear elastic

fracture mechanics was introduced by Kaplan in 1961 to determine fracture toughness of concrete

Numerous fracture tests were performed since then to measure the fracture behavior of concrete

[24-25] However, as research progressed, fracture process zone (FPZ) was found at the tip of a

crack in quasi-brittle materials, which causes the size effect of fracture parameters [26] Therefore,

several nonlinear fracture models were developed to measure fracture toughness, such as fictitious

crack model, crack band model, two-parameter fracture model, size effect model, effective crack

4

model and double-K fracture criterion (DKFC) [24-26]

These nonlinear fracture models can not only be applied to quasi-brittle materials, but can also be used in fiber-reinforced cementitious composites (FRCC) Ghasemi et al [1] obtained fracture toughness of FRSCC using the size effect model Kazemi et al [2] determined the fracture parameters of steel-fiber reinforced high strength concrete using the work of fracture method (WFM), which is based on the fictitious crack and size effect models Carpinteri et al [3] considered possible crack deflection during stable crack propagation and proposed a modified two-parameter fracture model to calculate Mode I plain-strain fracture toughness of polypropylene-fiber reinforced concrete Zhang et al [27] proposed a new crack extension resistance theory that simultaneously considers cohesive forces and bridging stresses to calculate the double-K fracture parameters of steel-fiber reinforced concrete Their results showed that unstable fracture toughness increased as the volume ratio of steel fiber increased However, initial fracture toughness did not have a similar trend

Fracture behavior based on 3-p-b test of cement-based materials reinforced by CaCO3whiskers and PVA-steel hybrid fibers did not receive a lot of attention in the literature Therefore, our goal is to study fracture behavior of CaCO3 whiskers-PVA-steel hybrid fibers reinforced cement mortar (MHFRC) CaCO3 whiskers, a new type of microfiber material, was mixed with PVA-steel hybrid fibers and then added into the cement mortar We studied how CaCO3 whisker

content, change of the PVA-steel fiber content and water cement ratios (w/c) affect mechanical propertied of MHFRC Fracture toughness (K IC ) and fracture energy (G F) were calculated based

on DKFC and WFM, respectively The results were compared to those obtained for unmodified cement mortar The synergy between CaCO3 whiskers, PVA and steel fibers was evaluated

quantitatively Moreover, relationship between the comprehensive reinforcing index (RI v ), w/c and

fracture parameters was developed This paper also describes crack resistance process and morphological characteristics of CaCO3 whisker, PVA and steel fibers in cement mortar and clarifies the microscopic reinforcing mechanism of the hybrid fibers Finally, empirical formulas

taking into account factors associated with both fiber (RI v ) and matrix (w/c) were proposed to

calculate fracture parameters of MHFRC

2 Fracture parameters

2.1 Double-K fracture criterion (DKFC)

Fracture toughness (K IC) is an important parameter in fracture analysis, and different fracture models were proposed for their determination fracture parameters Xu and Reinhardt [28-30] studied crack propagation of quasi-brittle materials and proposed analytical methods to determine fracture parameters A simplified method to determine double-K fracture parameters of three-point bending (3-p-b) tests was further developed [31] Two important characteristics, initial fracture

5

toughness (K ini

IC ) and unstable fracture toughness (K un

IC), were introduced to evaluate fracture behavior and to divide the crack propagation process into the three stages:

ini IC

K K , crack initiation stage

K K K , stable crack propagation stage (1),

un IC

K K , unstable crack propagation stage

where K is the stress intensity factor, which represents the crack-extension resistance under conditions of crack-tip plane strain in Mode I [32], K ini

IC is the initial fracture toughness, which represent the stress intensity factor at the initial crack tip when external load reaches the initial

cracking load, K un

IC is the unstable fracture toughness, which is defined as a stress intensity factor

at the critical effective crack tip at the external load equal to the peak load [26] According to

where P 0 is the initial cracking load, which can be obtained by strain gauge technique; S, W and B

are span, height and thickness of the beams, respectively (at S/W =4), a0 is length of the initial

crack and V 0 represents height ratios of beams, which can be obtained as a0/W

32.6 P (6),

where P max is the peak load; ac is the critical effective crack length, which corresponds to the peak

load; CMOD c is the critical crack mouth opening displacement, which corresponds to the peak

load; V c represents the height ratios of the beams which can be calculated as ac/W; H0 is the

thickness of the knife edges used to fix extensometer, and E is the Young’s modulus, which

obtained from the compressive cylinder measurements

2.2 Work fracture method (WFM)

Fracture energy (G F) is another important parameter in fracture analysis The work of fracture

6

method (WFM), which based on fictitious crack model, is the simplest and most popular method

to acquire the fracture energy [2] In our work, 3-p-b test of notched beams recommend by

RILEM 50-FMC was employed to determine the fracture energy [33] According to the WFM,

fracture energy can be calculated from the following equations:

G =(W mg ) / A (7),

0 0 0

where W 0 is the work of the external load and can be calculated by the area under the load-net

deflection, m is the weight of the beams, g represents acceleration due to gravity equal to 9.81m/s2,

δ 0 is the net deflection at the final failure of the beam, and A lig is area of the ligaments of notched

subsequent tests, their chemical compositions are shown in Table 2 Silica sand with a fineness

modulus of 1.9 and density of 2.65 g/cm3 was used as fine aggregate Superplasticizer from Sika

was used to ensure workability of the fresh mixture CaCO3 whiskers were provided by Youxing

Technology Co (Changde,China), and their chemical composition is also shown in Table 2

Smooth and straight PVA fibers were acquired from Wanwei High-Tech Material Co (Chaohu,

China) Steel fibers were from Bekaert OL Their properties and morphologies are presented in

Table 3 and Figure 1, respectively Defoaming agent, tributyl phosphate, was applied to eliminate

the bubbles in the mixture

Table 2 Chemical compositions of cement, fly ash and CaCO 3 whiskers wt.%

Table 3 Fibers properties

groups of MHFRCs with different w/c ratios were also used to study fracture parameters A constant sand cement ratio (s/c =0.5) was used in all groups The mix proportion of hybrid fiber and w/c of

the matrix is presented in Table 4

Table 4 Mix proportion of hybrid fiber and water cement ratio of matrix

8

Group

Water cement ratio

(w/c)

Steel fiber

PVA fiber

CaCO 3 whisker

Steel fiber

PVA fiber

CaCO 3 whisker

3.2 Samples preparation and experimental procedures

Raw materials were blended using UJZ-15 mortar mixer To ensure uniformity, dry raw materials, such as cement, fly ash, CaCO3 whiskers and silica sand, were mixed together for 120 seconds, after which water and superplasticizer were added, and the mixture was blended for another 120 seconds Constant flowability without segregation was used to control superplasticizer amount When cement mortar showed a good workability, PVA and steel fibers were added gradually and mixed to ensure their adequate dispersion Finally, a defoaming agent was added dropwise into the fresh mixture, which was then mixed for another 30 seconds to eliminate bubbles caused by the fiber addition The fresh mixture was carefully placed into special steel molds with 1mm thin blades at the mid-span of their side panels (40 × 40 × 200 mm in size) Schematic of the mold used in this work is illustrated in Figure 2 All samples were covered with a plastic sheet, and stored at laboratory temperature and demolded after 24 hours A precast crack was created at the mid-span of each beam during the casting process, after which these notched beams were cured for

28 days at 20 ± 2℃ and at over 95% of relative humidity according to GB/T 50081-2002 standard

9

[34] The thin blade fixed at the mid-span of the side panel of the steel mold could potentially influence fiber distribution and orientation at the tip of the thin blade, but it could not affect the fiber flow on the other side of the thin blade at the cross section at the mid-span In addition, this molding method satisfied the side loading requirement of side loading of flexural testing which described in ASTM C1609/C1609M-12

Figure 2 Schematic of the mold

In comparison to other fracture tests, 3-p-b tests were the simplest and had the minimum requirements for specimen forming and testing instruments Therefore, to determine the fracture parameters of different groups, 3-p-b fracture tests were performed on the notched beams by DKFC and WFM Each group in Table 3 had four same notched beams 40 × 40 × 200 mm in size Crack depth of each beam was16 mm

During the testing, an extensometer was fixed on a pair of knife edges to measure the crack mouth opening displacement A linear variable displacement transducer (LVDT) was used to measure the mid-span deflections of the notched beams The electric universal testing machine (WDW-50, Sinter, Changchun, China) was employed for 3-p-b fracture testing Displacement control was selected for loading, rate for which was maintained at 0.1 mm/min [2] Testing data were collected using DH3820 high speed static strain test analysis system at 10 Hz The span length

was 160 mm and the span depth ratio (S/W) was 4:1 The loading diagram of the notched beams is

shown in Figure 3

10

Figure 3 Loading diagram of the notched beams

A strain gauge (model: BX120-10AA; resistance: 120 Ω; sensitivity coefficient: 2.0; provided

by Zhejiang Huangyan Testing Apparatus Factory, China,) was employed to determine the initial cracking load, which is similar to the test method of the initial crack of concrete in the literature [28]

In this work, four strain gauges were arranged symmetrically on both sides of the beams to monitor their strain changes in the crack tip area to determine the crack load The locations of the strain

gauges are shown in Figure 3 When the applied load (P) was small, only elastic deformation existed

at the initial crack tip area of the notched beams, which did not crack However, as the applied load increased, the notched beams cracked because of the stress concentration at the crack tip Thus, the strain energy on both sides of the crack tip was released and strain retraction occurred All situations can be obtained by changing of the strain, and they are reflected in the corresponding P-ε curves In other words, strain reached maximum value and then returned, forming an inflection point The inflection point is the initial cracking point Initial cracking load P0 can be obtained from the P-ε curve (see Figure 4) The above process corresponds to the crack initiation stage which described

in the section 2.1, i.e K≤ K ini

IC Then, when P was greater than P 0, the fracture process zone was appeared at the tip of notch and the crack developed into the stable crack propagation stage that

described in the section 2.1, i.e K ini

IC ＜K ≤K un

IC With the further increase of P (when P＞P max ), the

development of crack entered into the unstable crack propagation stage, i.e K＞K un

IC

11

0.0 0.3 0.6 0.9 1.2 1.5 1.8

Uns

table cra

ck propagation stage

Crack

initiation stag

Figure 4 Typical P-ε curve

Basic mechanical properties of groups, such as modulus of elasticity, splitting tensile strength

and compressive strength, were also measured Three cylindrical samples φ100×200 mm in size were used to measure modulus of elasticity (E c) according to ASTM C469 standard [35] Splitting

tensile strength (f ft) tests were performed using three cubes with 70.7 mm sides GB/T 50081- 2002 standard [34] For each groups, six cubes were randomly selected from the notched beams after

fracture testing to evaluate their compressive strength (f cu) according to GB/T 17671- 1999 standard [36] Average values for all tests are presented in Table 5

Table 5 The basic mechanical properties

12

4 Results and Analysis

4.1 Fracture toughness measurements

K ini

IC and K un

IC values were calculated using experimentally obtained values and formulas (2) ~ (3) and (4) ~ (6), respectively Table 6 shows average values of fracture toughness of the notched beams Figure 6 presents initial fracture toughness values for different groups Addition of fibers

increased K ini

IC values, especially for the MHFRC group In comparison to the plain group, K ini

IC of the single-fiber reinforced cement mortar (sample CW10, P05 and S15) increased by 5.1, 13.7 and 154.5%, respectively Figure 6 (a) shows that as CaCO3 whiskers content increased, K ini

IC values first increased and then decreased Thus, even 1.0 vol % of CaCO3 whiskers content improved Kini

IC

values of PVA-steel hybrid fiber reinforced cement mortar However, excess of CaCO3 whiskers

resulted in the K ini

IC value decrease Figure 6 (b) shows how change of PVA-steel hybrid fiber

content affected on K ini

IC values We believe that an optimal combination of steel and PVA fiber contents mixed with 1.0 vol % of CaCO3 whiskers exist The above results showed that the proportion of 1.0 vol % CaCO3 whiskers, 0.5 vol % PVA and 1.5 vol % steel fibers was the best fiber

combination about K ini

IC Comparing to the plain group, K ini

IC value of S15P05W10 increased by

199.6% Ratio w/c greatly affected Kini

IC values (see Figure 6(c)): K ini

IC of S15P05W10 gradually

decreased as w/c increased S15P05W10a group demonstrated the highest K ini

IC equal to 1.164 MPa·m1/2 In comparison to the plain group, K ini

IC value for S15P05W10a increased by 320.2%

Table 6 The average values of fracture parameters

Group

K ini IC

(MPa·m1/2)

C1

Δa c (mm)

K un IC

(MPa·m1/2)

C2

G F (N/m)

0.2 0.4 0.6 0.8 1.0 1.2

W10bS15P05W S15P05

W10cS15P05

W10dS15P05

W10e

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

(c) Figure 6 Initial fracture toughness for different groups

FPZ also existed at the tip of the FRCC precast crack (see Table 6) similar to quasi-brittle materials FPZ in FRCC group is larger than that of plain This can be attributed to that the FPZ in FRCC not only exists cohesive force of matrix, but also has bonding force at the fiber-matrix

interface The Δa c in S15 showed the maximum value among the groups with single-fiber

Comparing to the plain group, Δa c of CW10, P05 and S15 improved 1.576, 1.189 and 3.616 times Figure 7 shows unstable fracture toughness for different groups Similar to the result of the

initial fracture toughness, addition of fibers significantly improved K un

IC Comparing to the plain

group, K un

IC value of CW10, P05 and S15 increased by 6, 15 and 944.7%, respectively Apparently,

14

S15 demonstrated greater contribution to K un

IC than that of CW10 and P05 MHFRC group (samples S15P05W00, S15P05W10, S15P05W20 and S15P05W30) showed higher Kun

IC values than the groups containing single-fiber reinforced cement mortar (with the exception of

S15P05W30 group, which K un

IC value lower than that of S15) (see Figure 7 (a)) As the CaCO3

whisker content in the MHFRCC group increased, K un

IC values first increased but then decreased

The reason for the initial increase of K un

IC is very likely presence of CaCO3 whiskers, which inhibit generation, development and penetration of micro-cracks by bridging effect On the other hand, micron-size CaCO3 whiskers can also fill the matrix pores and increase matrix compaction, thus,

improving the interfacial transition zone (ITZ) Decrease of K un

IC could be caused by poor dispersion of some of the CaCO3 whiskers added in excess: they agglomerate and cause matrix defects Change of PVA-steel hybrid fiber content for S10P10W10 and S12P08W10 groups did

affect K un

IC value (see Figure 7(b)), however, S15P05W10 and S18P02W10 groups behaved

differently Comparing to the S10P10W10, S12P08W10 and S15P05W10 groups, K un

IC value of S18P02W10 decreased by 51.2, 44.8 and 70.6%, respectively very likely because 0.2% of PVA

fiber is not high enough to play a significant role in the hybrid fiber system K un

IC value shown in

Figure 7 (c) show similar trends as K ini

IC value shown in Figure 6 (c) As w/c increased, K un

IC values

of S15P05W10 gradually decreased but not as steep (see Figure 7 (c)) This indicates that the

influencing factor of w/c is not sensitive to K un

IC, but content and type of fibers demonstrated much

greater influence on K un

IC values

CW10 P05 S15

S15P05W S15P05W S15P05W S15P05W

2 4 6 8 10 12

15

S15P05

W10aS15P05

W10bS15P05

W10S15P05

W10cS15P05

W10dS15P05

W10e

0 2 4 6 8 10 12 14

4.2 Fracture energy

Fracture energy is energy required to generate a crack with a certain unit area; it can be obtained by the area under the load-net deflection curve using 3-p-b tests on notched beams

Average values of G F of notched beams in different groups are presented in Table 6 Fiber addition

significantly improved G F values (especially for groups containing steel fibers) probably because the bridging action of macroscale steel fibers in post-cracking eliminated brittle failure and

enhanced energy dissipation capacity Values of G F of MHFRC increased with the addition of CaCO3 whiskers and PVA fibers (see Figure 8(a)) Comparing to the S15P05W00 group, fracture energy of groups containing CaCO3 whiskers demonstrated no significant improvements very likely because of the micron-size CaCO3 whiskers, which only arrest cracks on the microscale and play a

role in pro-cracking stage As steel fiber content increased and PVA fiber content decreased, G F

value first increased but then decreased (see Figure 8 (b)) Comparing to S10P10W10,

S12P08W10 and S18P02W10, G F value for S15P05W10 increased by 55.1, 37.2 and 31.7%,

respectively Fracture energy of S15P05W10 decreased significantly as w/c changed from 0.24 to

0.40 (see Figure 8 (c)) probably because of interface bond strength effect between the fibers and the

matrix Higher w/c in S15P05W10 leads to higher porosity, which can weaken the bonding strength

at the fiber-matrix interface G F value also declined sharply at w/c > 0.30 Comparing to

S15P05W10, fracture energy of S15P05W10c, S15P05W10d and S15P05W10e decreased by 43.0, 106.5 and 150.0%, respectively