- Email: [email protected]
The impact resistance and mechanical properties of reinforced self-compacting concrete with recycled glass fibre reinforced polymers
Accepted Manuscript The impact resistance and mechanical properties of reinforced self-compacting concrete with recycled glass fiber reinforced polymers M. Mastali, A. Dalvand, Assistant Professor, A. Sattarifard PII:
S0959-6526(16)30069-5
DOI:
10.1016/j.jclepro.2016.02.148
Reference:
JCLP 6858
To appear in:
Journal of Cleaner Production
Received Date: 4 November 2015 Revised Date:
23 February 2016
Accepted Date: 26 February 2016
Please cite this article as: Mastali M, Dalvand A, Sattarifard A, The impact resistance and mechanical properties of reinforced self-compacting concrete with recycled glass fiber reinforced polymers, Journal of Cleaner Production (2016), doi: 10.1016/j.jclepro.2016.02.148. 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, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
The impact resistance and mechanical properties of reinforced selfcompacting concrete with recycled glass fiber reinforced polymers
RI PT
M. Mastali1, A. Dalvand2*, A. Sattarifard 3 1. ISISE, Minho University, Department of Civil Engineering, Campus de Azurem, Guimaraes, 4800-058 Portugal
SC
2. Assistant Professor, Department of Engineering, Lorestan University, Khorramabad, Iran
M AN U
Corresponding author Email: [email protected]
3. Faculty of Civil Engineering, Semnan University, Semnan, Iran
Abstract
TE D
Experimental and statistical analyses are represented considering the impact resistance and mechanical properties of self-compacting concrete reinforced with recycled Glass Fiber Reinforced Polymers (GFRP). Specimens were reinforced with recycled GFRP in three
EP
groups, including 0.25%, 0.75%, and 1.25% of fiber volume fractions. An extensive experimental program including two hundred and fifty two specimens were prepared and
AC C
tested to characterize the impact resistance and mechanical properties of the reinforced selfcompacting concrete with glass fiber reinforced polymers. Scanning Electron Microscope images were used to perceive the failure mechanism of recycled glass fiber in the matrix. Then, the relatively large and reliable collected data from the experimental tests was used to start statistical and analytical analysis. The results showed that adding recycled glass fiber reinforced polymers results in improving the impact resistance and mechanical properties of the reinforced self-compacting concrete with glass fiber reinforced polymers. Moreover, the statistical data analysis revealed that the
ACCEPTED MANUSCRIPT impact resistance and mechanical properties of reinforced self-compacting concrete follow a normal distribution. Furthermore, some equations were developed to correlate the impact resistance of the reinforced self-compacting specimens to the mechanical properties with high values of coefficient of determination.
RI PT
Keywords: Recycled GFRP fiber; Self-compacting concrete; Mechanical properties; Impact
AC C
EP
TE D
M AN U
SC
resistance.
ACCEPTED MANUSCRIPT
1. Introduction Using some recyclable waste materials in different products is favourable because it is economic and environmental-friendly and therefore, researchers have started to use recycled materials in the concrete mixes, which is one of the main construction materials.
RI PT
Significant studies have been conducted to use waste materials in concrete using different ways such as reinforcement of concrete with recycled fibers or using waste materials as powder to replace Portland cement (Jianzhuang X. et al., 2015; Abdollahnejad Z. et al.,
SC
2015; Ghasemi N.M. et al., 2014; Foti D., 2013)
Foti D. in 2011 worked on the reinforced concrete with Polyethylene Terephthalate (PET)
M AN U
bottles waste fibers. She found that adding small amounts of recycled fibers from PET bottle wastes could have significant effects on improvement of post-cracking performance of compositions. Moreover, it was revealed that these fibers enhanced the toughness of specimens. In 2013, Foti D. also studied the feasibility of recycling PET fibers, obtained from
could increase ductility.
TE D
waste bottles with different shapes. Her studies revealed that using PET fibers in concrete
The applicability of self-compacting concrete containing waste Polyethylene Terephthalate (PET) particles was assessed using slump flow, V-funnel and L-box tests. Moreover,
EP
mechanical properties of mixtures including compressive, tensile and flexural strengths, and
AC C
modulus of elasticity were evaluated. The experimental results indicated that waste PET particles could be reused as aggregates in self-compacting concrete, while using the waste PET particles led to degrading mechanical properties. However, fly ash and silica fume compensated the loss of strength caused by adding PET (Sadrmomtazi A. et al., 2016). Using Fiber Reinforced Polymer (FRP) materials has increased significantly in several applications, including industry, construction, and transportation due to its light weight, high stiffness, high tensile strength, good durability resistance, and high temperature resistance. In this regard, many products are recently manufactured by FRP materials in the construction industry such as sheets for strengthening structural elements (like beams,
ACCEPTED MANUSCRIPT columns, and slabs), bars to reinforce concrete due to good properties of FRP materials against corrosion, structural profiles, and sandwich panels (Mastali M. et al., 2015). Production process usually results in some waste FRP materials, which should be recycled through material recycling, chemical recycling, or thermal recycling. Otherwise, they are
RI PT
presumed to be buried in the ground or remain unusable. Asokan P. et al. investigated the recycling potential of glass fibre reinforced plastic waste in concrete and cement composites. Results revealed that the compressive strength of specimens using 5%–50% Glass Fibre Reinforced Plastic (GFRP) waste powder under
SC
water curing varied from 19 MPa to 37 MPa. Increasing the content of GRP waste decreased the compressive strength. The density of concrete with 50% GRP waste reduced
M AN U
about 12% compared to the reference concrete. Moreover, the flexural strength for specimens using 5% GRP waste fibre was recorded as 16.5 MPa, while the flexural strength for control beam was measured as 10.5 MPa.
Keilji et al. in 2005 studied mechanical properties of concrete reinforced with Carbon Fiber
TE D
Reinforced Polymer (CFRP) fibers. The applied CFRP fibers were provided from waste CFRP sheets. In their study, they investigated the effects of size and content of chopped CFRP fibers on the compressive and flexural strengths. They indicated that the improvement
EP
of mechanical properties of reinforced concrete with CFRP fibers is governed significantly by the size of fibers. In 2014, García D. et al. worked on evaluation of mechanical properties of
AC C
recycled Glass Fiber Reinforced Plastics (rGFRP) from different sources as short fibers. The physical–mechanical properties of the concrete reinforced with rGFRP were characterized by conducting compressive and flexural tests as well as shrinkage stability and alkaliaggregate reaction. Moreover, the interaction between the rGFRP and the matrix was observed and characterized by Scanning Electron Microscope (SEM) images. They found that the compressive and flexural strength increased 22% and 16% respectively compared to control specimens, after providing rGFRP materials from an optimized milling process (García D. et al., 2014).
ACCEPTED MANUSCRIPT Successful application of recycled fibers to reinforce plain concrete enhances the impact resistance, and improves mechanical and durability properties. However, this improvement strongly depends on the composition, and also mechanical properties and contents of recycled fibers. Therefore, some studies have been performed on mechanical properties of
RI PT
reinforced concrete with some types and contents of recycled fibers under static loading. Since the impact resistance and mechanical properties of reinforced self-compacting concrete with recycled fibers are not characterized completely yet, in the present paper, an experimental/statistical study was conducted to characterize the impact resistance and
SC
mechanical properties of self-compacting concrete reinforced with recycled glass fiber in different volume fractions, including 0.25%, 0.75%, and 1.25%. Considering statistical
M AN U
analysis is appropriate for better understanding the impact resistance and mechanical properties of reinforced self-compacting concrete. To the best of the author’s knowledge, an extensive experimental/statistical study on impact resistance and mechanical properties of reinforced self-compacting concrete made with rGFRP fibers has not been reported yet.
TE D
Although some studies have considered mechanical properties of reinforced normal concrete with recycled GFRP fibers (García et al., 2014). Yet, they have not included a statistical assessment based on a relatively large number of specimens under compression,
AC C
EP
flexure, and impact test.
2. Experimental program 2.1.
Materials and concrete mix design
In this study, mixture compositions are a combination of Portland cement 42.5R, fine aggregates, silica fume, water, and superplasticizer (SP). Table 1 represents the chemical and physical properties of Portland cement 42,5R. Also, the proportions of materials used in the mixtures are mentioned in Table 2. The contents of materials in Table 2 have been obtained according to a target slump flow equal or more than 600 mm for plain self-
ACCEPTED MANUSCRIPT compacting concrete. In this regard, some preliminary experimental tests were implemented on workability of some plain self-compacting concretes. The results revealed that using cement to fine aggregate ratios equal to one resulted in slump flows of lower than 600 mm.
Table 1.
RI PT
Table 2. The particle distribution of fine aggregate is shown in Fig. 1. The workability of mixtures was adjusted through a high range water reducer agent with the commercial name of Mape110.
SC
Moreover, GFRP fibers were recycled from the remained unusable GFRP sheets in the Civil Lab of Semnan University. In this regards, GFRP sheets were cut to smaller GFRP fibers
M AN U
with an average length of 20 mm using a shred machine, as shown in Fig. 2. The thickness of GFRP sheets was measured as 0.11 mm. The tensile strength, elastic modulus, and density of the GFRP sheets, used to produce the GFRP fibers, were 1103 MPa, 44.8 GPa, and 2080 kg/m3, respectively. All efforts were made to provide equal recycled glass fiber lengths.
TE D
Fig 1. Fig 2.
To batch, cement was mixed with fine aggregate for one minute. Then, the water and super
EP
plasticizer were added to the composition that was mixed for 6–8 min. Finally, the concrete-
AC C
fibers mixtures were prepared by inchmeal addition of the glass fibers to the fresh selfcompacting concrete while adding enough fibers to concrete mixture so that the desired ratio of fibers to volume was achieved. Chopped glass fibers were incrementally added to the self-compacting concrete to avoid unfavourable effects such as balling fibers. Moreover, the homogeneity of the composition was assessed by analysing the microstructure of mixtures through microscopic images. This analysis was executed on the hardened reinforced concrete. Then, mixed self-compacting concrete was cast into cubic molds (100×100×100 mm), cylindrical disk molds (150×65 mm), and prismatic beams (500×60×60 mm) to be tested in compressive, impact, and flexural tests, respectively. The specimens were sealed
ACCEPTED MANUSCRIPT and retained at 250C and 90% relative humidity for 24 hours. After demolding, specimens were cured in water at 230C for 28 days and all specimens were tested afterwards. The fresh state behavior of plain self-compacting concrete was assessed using the slump test and the V funnel test (EN 12350-8:2010, 2010; EN 12350-9: 2010, 2010; Shahid Iqbal et
Fig 3.
Test setups and test procedures
2.2.1.
Compressive test
SC
2.2.
RI PT
al., 2015). Fig. 3a presents the slump flow test for reinforced self-compacting concrete.
According to the compositions listed in Table 2, sixty-four cubic specimens with dimensions
M AN U
of 100 ×100 × 100 mm were cast and tested in three different reinforced groups to study the effects of different recycled glass fibers volumes on the compressive strength, including 0.25%, 0.75%, and 1.25% recycled glass fibers. Four cubic specimens were used to evaluate the compressive strength of plain self-compacting concrete and twenty cubic specimens were employed to assess the compressive strength of each group of reinforced
2.2.2. Flexural test
TE D
specimens with rGFRP fibers.
EP
To assess the flexural performance of reinforced self-compacting concrete with recycled
AC C
glass fibers, sixty-four prismatic beams (60 × 60 × 500 mm) were cast and tested under Four Point Bending (FPB) test. The test setup used to perform flexural assessment is depicted in Fig. 4a.
Fig 4.
The flexural loading was applied to the beams with a loading rate of 0.5 mm/min. Moreover, a Linear Variable Differential Transformer (LVDT) of 10 mm stroke was used to record midspan deflection. As mentioned before, fewer variables are involved in mechanical properties of plain self-compacting concrete compared to reinforced self-compacting concrete. In this regard, four prismatic specimens were used to assess the flexural strength of plain self-
ACCEPTED MANUSCRIPT compacting concrete, while twenty prismatic specimens were required for each group of prismatic beams reinforced with rGFRP fibers.
2.2.3. Impact drop weight test
RI PT
The drop weight impact test was conducted based on ACI Committee 544 on fiber reinforced concrete. The apparatus is depicted in Fig. 4b (ACI Committee 544, 1988), in which a steel hammer of 4.45 kg drops from a height of 457 mm on a steel ball with a diameter of 63.5
SC
mm. This steel ball is in contact with central surface of specimens. Totally, one hundred and twenty four disks with a diameter of 150 mm and a height of 65 mm were cast and prepared
M AN U
to test at the age of 28 days. To calculate energy absorption, the following equation is used: Impact energy (En)= N × W × H
(1)
where, N is number of blows, W is weight of steel hammer with a mass of 4.45 kg, and H is the height of fall. To assess the impact resistance of disks under drop weight impact load, four cylinders were used to evaluate the impact resistance of plain concrete, while for each
TE D
group of reinforced cylinder with rGFRP fibers, forty cylinders were used.
3.1.
EP
3. Results and discussion Fresh mix properties
AC C
The homogeneity of mixtures was investigated by analysing the microstructure of the prepared mixes. Fig. 3b shows that the rGFRP fibers are distributed homogeneously regardless of the fiber content, considerable amounts of the rGFRP fibers well attached to the cement-based matrix.
Fig. 5 presents the results recorded for the fresh state properties of concrete mixtures. Fig. 5a shows that increasing rGFRP fiber results in reduction of the slump flow, so that the maximum reduction in slump flow is recorded as 15% for specimen GFRC1.25. Moreover, it was found that Tv and T500 increased linearly as the fiber content increased.
ACCEPTED MANUSCRIPT Fig 5. Preliminary experimental tests were also carried out on the workability of the reinforced selfcompacting concrete with 2% recycled GFRP fiber. The obtained results for the reinforced self-compacting concrete with 2% recycled GFRP fiber showed that the workability reduced
RI PT
significantly compared to the reinforced self-compacting concrete with 1.25% rGFRP fiber. Therefore, the study focused on the reinforced self-compacting concrete specimens with
SC
GFRP fiber content of lower than 2%.
3.2.1. Compressive strength
M AN U
3.2. Hardened concrete properties
As previously mentioned, the evaluation of compressive strength of cubic specimens was carried out using sixty-four cubic specimens of 100 × 100 × 100 mm. A digital standard
TE D
automatic testing machine with a capacity of 600 kN was used to apply compressive loading. The compressive strength of specimens was measured and recorded in Table 3. Table 3.
EP
The compressive strength of specimens shows that reinforcement of plain self-compacting
AC C
concrete results in improving the compressive strength, so that this improvement is significantly governed by recycled glass fiber contents. Adding glass fiber leads to bridging action in the reinforced cement-based composites. Transferring stress from glass fibers to the matrix regard to interfacial shear strength offers more resistance to crack opening in the specimens. As shown in Fig. 6a, reinforcement of plain self-compacting concrete resulted in an increase of 25.52%, 39.57%, and 47.92% in compressive strength of specimens with recycled glass fiber of 0.25%, 0.75%, and 1.25%, respectively. This improvement could be attributed to arresting cracks in specimens by glass fibers under compressive loading. Fig 6.
ACCEPTED MANUSCRIPT Fig 9a represents the histogram of compressive strength of specimen GFRC0.25. The indicated results depicts that the compressive strengths of specimens are normally distributed and fit well with the superimposed normal distribution curve (more discussions on statistical analysis will be presented in section 4.2). GFRC1.25 had the highest mean
RI PT
compressive strength (59.17 MPa) among all the groups, while the highest coefficient of variation (COV) of 8.38% belonged to this group. Additionally, as it could be seen in Table 3, increasing glass fiber content results in increased scatter of compressive strength.
SC
3.2.2. Flexural strength
M AN U
Sixty-four prismatic beams of 500 × 60 × 60 mm were employed to be assessed under Four Point Bending (FPB) test. Table 4 indicates the registered flexural strengths of prismatic beams. Considering the recorded results, increasing glass fiber content leads to enhanced flexural performance and increased flexural strength of beams. Fig. 6b presents the efficiency of recycled glass fiber on increasing the flexural strength of specimens.
TE D
Reinforcement of plain self-compacting concrete beams with recycled fiber volume fractions of 0.25%, 0.75%, and 1.25% led to 30.13%, 42.93%, and 59.46% increase in the flexural strength of specimens, respectively (see Fig. 6b). Additionally, the maximum flexural
EP
strength was measured as 5.98 MPa for specimen GRPC1.25, which is 22.54% and 11.56% more than mean flexural strengths of specimens with 0.25% and 0.75% recycled GFRP
AC C
fibers, respectively. Moreover, it was observed that like compressive strength, increasing the content of recycled glass fiber leads to increased scatter of the results obtained for flexural strengths, so that the maximum COV was recorded as 9.61% for specimens reinforced with 1.25% of glass fiber.
Table 4. Enhancement of flexural performance of reinforced beams with recycled GFRP fibers can be due to the bridging action of GFRP fiber, through which the bridging fibers can partially transfer the stress across the crack. In 2015, Felekoglu et al. found that increasing
ACCEPTED MANUSCRIPT homogeneous distribution of fibers throughout the composite may also increase the number of formed cracks. The frequency histogram of flexural strength of specimen GFRC0.25 is indicated in Fig 9b. This figure shows that flexural strengths follow a normal distribution.
RI PT
3.2.3. Impact resistance The drop weight impact test was conducted on one hundred twenty four cylinder disks. The recorded impact results are presented in Table 5. Reinforcement of plain self-compacting
SC
concrete with recycled glass fiber of 0.25%, 0.75%, and 1.25% increased the first crack impact resistance of disks 2.53, 3.73, and 5.06 times, respectively. Furthermore, as shown in
M AN U
Figs 6c and 6d, the ultimate crack impact resistance increased 2.94, 4.44, 6.14 times for reinforced cylinders with recycled glass fiber volume fractions of 0.25%, 0.75%, and 1.25%, respectively. Like the results attained for mechanical properties of reinforced specimens, reinforcement of disks resulted in increasing the impact resistance, so that both the first crack impact resistance and the ultimate crack impact resistance increased significantly.
TE D
Table 5.
As listed in Table 5, the maximum value of the first crack impact resistance and the ultimate impact resistance were recorded 75.97 and 98.32 blows for reinforced disks with 1.25%
EP
glass fiber, respectively.
By comparing the recorded first and ultimate crack impact resistance of specimens, it is
AC C
revealed that the first and ultimate crack impact resistance of disks reinforced with 1.25% of glass fiber increase 35.73% and 38.38% as compared to the first and ultimate crack impact resistance of reinforced disks with 0.75% of recycled glass fiber, respectively. Moreover, this increase is 99.55% and 2.08 times for specimens reinforced with 0.25% fiber. As it occurred for mechanical properties of specimens, increasing glass fiber content results in increased scatter of the first and ultimate impact resistance, so that the maximum COV for the first and ultimate crack resistance were recorded as 43.41% and 43.49% for specimens of GFRC1.25, respectively.
ACCEPTED MANUSCRIPT ACI Committee 544’s repeated drop-weight impact test for concrete is often criticized for large variations within the results, therefore, a modified impact test was proposed to ACI test thatsignificantly improved the reliability of the results (Badr et al., 2005). The impact test in the present paper was implemented based on ACI Committee 544, therefore, the recorded
RI PT
large scatter can be derived from large variations involved in the repeated drop-weight impact test.
Table 6 presents some statistical results from experimental data. It shows that increasing fiber content results in increasing the interval difference between the minimum and
resistance of specimens.
M AN U
Table 6.
SC
maximum values of compressive and flexural strengths, and also the first/ultimate impact
The number of blows to reach failure impact resistance of disks varied from 38.07 to 47.07 for the specimens GFRC0.25, from 55.97 to 71.05 for the specimens of GFRC0.75, and from 75.97 to 98.32 for the specimens of GFRC1.25. Using more recycled glass fiber increased
TE D
the difference between blow number of the first crack impact resistance and the blow number of the ultimate crack impact resistance. Here in this study, increasing the Number of Post initial crack Blows to failure is called ‘‘INPB’’ parameter. Increasing glass fiber content
EP
from 0.25 to 0.75% and 1.25% increased the number of post-initial crack blows to failure 67.55% and 2.48 times, respectively. Moreover, reinforcement of plain self-compacting
AC C
concrete with 0.25%, 0.75%, and 1.25% increased the number of post-initial crack blows to failure of 9, 15.08, and 22.35 times, respectively. Based on equation (1), the maximum absorbed energy for the first and ultimate impact resistances were calculated as 1518.92 J and 1965.00 J for specimens of GFRC1.25, respectively. The frequency histograms of the first and ultimate impact resistance of specimen GFRC0.25 are shown in Figs 9c and 9d, which indicate that the first and ultimate impact resistance of specimens hardly follow a normal distribution due to low p-values listed in Table 6. Fig. 7 depicts the formed crack patterns on some tested disks, which were reinforced with 0.25%, 0.75%, and 1.25% recycled glass fiber. Increasing fiber content resulted in more cracks on the surface of
ACCEPTED MANUSCRIPT specimens, as shown in Fig. 7. Forming multiple cracks is due to fiber bridging action of glass fibers. Additionally, a fractured surface is indicated in Fig. 7, where parts of the recycled glass fibers are marked. Respect to the presented fractured surface, it is revealed that recycled glass fibers are distributed uniformly in the section.
RI PT
Fig 7. SEM images were used to reach a profound understanding about the fiber reinforcement mechanisms. As the desired bonding was formed between glass fibers to cement based
SC
materials, cement-hydrated particles almost covered the surface of fibers, as shown in Fig. 8. The SEM images obtained in the fracture surface of tested specimens show a tendency
M AN U
for the rupture of the long glass fibers. Fig 8.
4.
Statistical and analytical analysis
TE D
In this section, probable distribution of compressive strength, flexural strength, and impact resistance of recycled glass fiber reinforced self-compacting concretes is discussed. Most statistical data analyses in the literature are related to mechanical properties of reinforced
EP
concrete made with the steel and Polypropylene (PP) fibers (Song P.S. et al., 2005a; Song P.S. et al. 2005b; Natarajaa M.C. et al., 1999; Fakharifar M. et al., 2014; Rahmani T. et al.,
AC C
2012) and no study has been conducted on mechanical properties of self-compacting concrete reinforced with recycled GFRP fibers so far. Analysing a large experimental database provided a reliable source for implementing statistical studies. Statistical analysis establishes a better understanding of mechanical properties of these materials. Effect of type and volume fraction of fiber on statistical parameters and distribution of flexural strength, compressive strength, and impact strength are important to be described. Moreover, some empirical relations are developed for mechanical properties of specimens, as well as compressive strength of specimens used to predict impact resistance of specimens.
ACCEPTED MANUSCRIPT
4.2.
Probability distribution
To conduct the statistical analysis, the OriginLab computer program was used to find the probability distributions of mechanical properties. To test the normality of data, Kolmogorov–
RI PT
Smirnov (KS test) and Anderson–Darling (AD test) methods were used. Kolmogorov– Smirnov test concentrates on the maximum deviation between the observed cumulative histogram and the hypothesized cumulative distribution function (Jack R. Benjamin et al.,
SC
1970). This feature of fit test evaluates the feasibility of using the normal distribution for the impact resistance and mechanical properties of fiber reinforced self-compacting concrete.
M AN U
For both methods, a p-value is computed to quantify the strength of the evidence against the null hypothesis (Rahmani T. et al., 2012). A p-value of smaller than 0.05 indicates that the null hypothesis (confirming normal distribution) is rejected. The hypotheses for the normal distribution are as follows:
Ho1. The compressive strength of recycled glass fibre reinforced self-compacting concrete
TE D
follows the normal distribution;
Ho2. The flexural strength of recycled glass fibre reinforced self-compacting concrete follows the normal distribution;
EP
Ho3. The first crack impact resistance of recycled glass fibre reinforced self-compacting concrete follows the normal distribution;
AC C
Ho4. The ultimate crack impact resistance of recycled glass fibre reinforced self-compacting concrete follows the normal distribution; Ha. Not Ho1, 2, 3, 4.
The obtained p-values for all tests are listed in Table 6, which are all above 0.05. This finding corroborates the null hypothesis at the 0.05 significance level. Therefore, the compressive strength, flexural strength, first crack impact resistance, and ultimate crack impact resistance of reinforced SCC with rGFRP follow the normal distribution. It is worth stating that mechanical properties of the specimens are better fitted to the normal distribution compared
ACCEPTED MANUSCRIPT to the first and the ultimate crack impact resistance of the specimens, due to higher p-values of the compressive strength and flexural strength compared to the p-values registered for the first crack impact resistance, ultimate impact resistance. This result confirms the findings of Fakharifar et al. (Fakharifar M. et al., 2014; Rahmani T. et al., 2012). Fig. 10 indicates the
RI PT
normal probability curves of the impact resistance and mechanical properties of GFRC specimens. As shown in Fig. 10, the nearly linear pattern of the data of compressive and flexural strengths implies that mechanical properties of GFRC specimens follow a normal
Fig 9.
M AN U
Fig 10.
SC
distribution.
The coefficients of variation of the specimens, tested under drop weight impact test, are listed in Table 5. Table 5 could be used to determine the minimum replication number of tests ‘‘n’’, required for ensuring that the error percentage ‘‘e’’ in the measured average value is below a certain limit. This number of tests can be calculated by equation (2) (Swamy R.N.
TE D
et al., 1976):
2 [ COV ] t 2 n=
e2
(2)
EP
Where, “t” in equation (2) is the value of Student t-distribution for a certain level of confidence, “e” is the error percentage (10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, and
AC C
50%), and COV is the coefficient of variation. The t-distribution is a probability distribution used to estimate population parameters when the sample size is small and/or when the population variance is unknown. The value of Student t-distribution depends on the level of confidence and the degree of freedom, which is related to the number of samples. The value of “t” for degrees of freedom of more than 120 can be considered as 1.645 and 1.282 at 95% and 90% level of confidence, respectively (Box GEP et al., 1978; Moore D.S. et al., 1989). Table 7 shows the minimum number of replications required to keep the error on the specified limits at 90% level of confidence. Table 7 also illustrates that if the error is lower than 10% for the first crack impact resistance, the minimum number of required replications
ACCEPTED MANUSCRIPT would be 23, 25, and 31 for specimens of GFRC0.25, GFRC0.75, and GFRC1.25, respectively at 90% level of confidence. Moreover, for the ultimate crack impact resistance, the minimum number of required replications would be 24, 28, and 32 for specimens of GFRC0.25, GFRC0.75, and GFRC1.25, respectively at 90% level of confidence. This table
RI PT
also indicates if the error remains under 10%, the minimum numbers of replications for GFRC0.25, GFRC0.75, and GFRC1.25 are 37, 41, and 51 for the first-crack strength and 40, 45 and 53 for the failure strength, respectively at 95% level of confidence. According to the
required at each level of error.
4.3.
M AN U
Table 7.
SC
results represented in Table 7, increasing fiber content increases the number of tests
Analytical study
According to the relatively large collected experimental database, the impact resistance and mechanical properties of reinforced concrete self-compacting specimens can be correlated through empirical relations with high coefficient of determination (R2). These empirical
TE D
relations were developed using regression analysis. Fig. 11a indicates the linear empirical equations that relate the flexural strength to the compressive strength. Based on the obtained equations, flexural strengths can be correlated
EP
linearly to compressive strengths for each group of reinforced self-compacting concrete specimens. The slope of developed equation for GFRC0.75 is higher than specimens of
AC C
GFRC0.25 and GFRC1.25. It should be noted that adding 0.75% glass fiber results in the highest rate of flexural strength. Fig 11.
Moreover, based on Fig. 11b, it was revealed that the highest rate of gaining the INPB is achieved for reinforced disks containing 0.75% recycled glass fiber. Fig. 12 depicts the correlation between the compressive strength and the first or ultimate crack impact resistance. The results demonstrate that there is a linear relationship between compressive strength and the first/ultimate crack impact resistance of specimens. Moreover, the ultimate
ACCEPTED MANUSCRIPT impact resistance of GFRC0.75 had the highest slope followed by GFRC1.25 and GFRC0.25. It means that specimens of GFRC0.75 have the best impact performance. Fig 12. Furthermore, the relation between the first impact resistance versus the ultimate crack
RI PT
resistance of specimens is shown in Fig. 13. A linear relation between the first and ultimate crack resistance of disks is depicted in Fig. 13 with coefficient of determination (R2) of more than 0.96. Linear relationships between the first and ultimate crack impact resistance is also reported (Fakharifar M. et al., 2014) for high performance fiber reinforced concrete with
SC
polypropylene (PP) fibers. According to the recorded results, after forming the first crack on
M AN U
the reinforced specimens, the best performance in increasing the ultimate crack impact resistance was detected in GFRC1.25 specimens. Fig 13.
5.
Conclusion
TE D
An experimental and analytical study was established to investigate the effects of adding recycled GFRP fibers on the impact resistance and mechanical properties of self-compacting concrete. Therefore, two hundred and fifty two specimens were tested experimentally under
EP
compressive test, flexural test, and drop weight impact test and the obtained experimental results were interpreted. Moreover, using SEM images, the failure mechanism of recycled
AC C
glass fiber in the matrix was characterized. Moreover, a statistical data analysis was carried out to better understand the impact resistance and mechanical properties of these materials. In this regard, the relatively large experimental database was used for statistical and analytical analysis and consequently the following results were concluded: 1. Increasing GFRP fiber content resulted in linearly reduced slump flow, while T500 and Tv increased linearly.
ACCEPTED MANUSCRIPT 2. Adding GFRP fiber improves mechanical properties and the impact resistance, so that the maximum compressive strength, flexural strength, and impact resistance were recorded for specimen GFRP1.25. 3. The maximum rate of gaining flexural strength was recorded for GFRC0.75 and the
RI PT
best performance in terms of the first crack and the ultimate impact resistance was recorded for specimens GFRC1.25 and GFRC0.75, respectively.
4. The highest rate of gaining the number of post initial crack blows to failure was obtained for GFRC0.75.
SC
5. Flexural strength, INPB, the first and ultimate impact resistance linearly correlated to compressive strength to the coefficient of determination (R2) of more than 0.86.
M AN U
6. The first and ultimate crack impact resistance are correlated linearly. 7. Increasing glass fiber leads to increased scatter of the results for mechanical properties and impact resistance that can affect fiber/matrix interfacial bond. 8. The compressive strength and flexural strength of the specimens follow the normal
TE D
distribution, while the first and the ultimate crack impact resistance hardly follow a
References
Abdollahnejad Z., Pacheco-Torgal F., Félix T., Tahri W., Barroso Aguiar J., (2015), “Mix
AC C
•
EP
normal distribution.
design, properties and cost analysis of fly ash-based geopolymer foam”, Journal of Construction and Building Materials, Vol. 80, pp: 18-30. •
ACI Committee 544, (1988), Measurement of properties of fiber reinforced concrete. ACI 555 Material Journal, Vol. 85, pp: 583–593.
•
Asokan P., Osmani M., Price A.D.F., (2009), “Assessing the recycling potential of glass fibre reinforced plastic waste in concrete and cement composites”, Journal of Cleaner Production, Vol. 17, pp: 821-829.
ACCEPTED MANUSCRIPT •
Badr A., Ashour F. Ashraf, (2005), “Modified ACI Drop-Weight Impact Test for Concrete”, ACI materials Journal, Vol. 102, pp: 249-255.
•
Box GEP, Hunter WG, Hunter J.S., (1978), “Statistics for experimenters”, Wiley, USA.
•
Burak Felekoglu, Kamile Tosun-Felekog, Eren Gödek, 2015, “A novel method for the
RI PT
determination of polymeric micro-fiber distribution of cementitious composites exhibiting multiple cracking behavior under tensile loading”, Journal of Construction and Building Materials, Vol. 86, pp: 85-94.
EN 12350-8:2010. Testing fresh concrete. Part 8: Self-compacting concrete. Slump-
SC
•
flow test. 2010.
EN 12350-9:2010. In: Testing fresh concrete. Part 9: Self-compacting concrete. V-
M AN U
•
funnel test. 2010. •
Fakharifar M., Dalvand A., Arezoumandi M., Sharbatdar M.K., Chen G., Kheyroddin A., (2014), Mechanical properties of high performance fibre reinforced cementitious composites, Journal of Construction and Building Materials, Vol. 71, pp: 510-520. Foti D., (2011), “Preliminary analysis of concrete reinforced with waste bottles PET
TE D
•
fibers”, Journal of Construction and Building Materials, Vol. 25, pp: 1906-1915. •
Foti D., (2013), “Use of recycled waste pet bottles fibers for the reinforcement of
•
EP
concrete”, Journal of Composite Structure, Vol. 96, pp: 396–404. García D., Vegas I., Cacho I., (2014), “Mechanical recycling of GFRP waste as short-
AC C
fiber reinforcements in microconcrete”, Journal of Construction and Building Materials, Vol. 64, pp: 293-300. •
Ghasemi, N.M., Sharbatdar, M.K., Mastali, M., (2014), “Repairing reinforced concrete slabs using composite layers”, Journal of Materials and Design, Vol. 58, pp: 136-144.
•
Jianzhuang X., Long Li, Luming Shen, Chi Sun Poon, (2015), “Compressive behaviour of recycled aggregate concrete under impact loading”, Journal of Cement and Concrete Research, Vol. 71, pp: 46-55.
ACCEPTED MANUSCRIPT •
Jack R Benjamin, C. Allin Cornell, (1970), Probability, Statistics, and Decision for Civil Engineers, Dover publication, New York, USA.
•
Keiji O., Tomoyuki S., Makoto M., (2005), “Strength in concrete reinforced with recycled CFRP pieces”, Journal of Composite Part A: Applied Science and Manufacturing, Vol.
•
RI PT
36, pp: 893-902. Mastali M., Valente I.B., Joaquim A.O. Barros, Delfina M.F. Gonçalves, (2015), “Development of innovative hybrid sandwich panel slabs: Experimental results”, Journal
•
SC
of Composite Structures, Vol. 133, pp: 476-498.
Moore D.S., McCabe G.P., (1989), “Introduction to the practice of statistics”, New York:
•
M AN U
W.H. Freeman & Company, USA.
Natarajaa M.C., Dhangb N., Guptab A.P., (1999), “Statistical variations in impact resistance of steel fiber-reinforced concrete subjected to drop weight test”, Journal of Concrete and Cement Research, Vol. 29, pp: 989-995.
•
Rahmani T., Kiani B., Shekarchi M., Safari A., (2012), Statistical and experimental
TE D
analysis on the behavior of fibre reinforced concretes subjected to drop weight test, Journal of Construction and Building Materials, Vol. 37, pp: 360-369. •
Sadrmomtazia A., Dolati-Milehsaraa S., Lotfi-Omrana O., Sadeghi-Nikb A., (2016),
EP
“The combined effects of waste Polyethylene Terephthalate (PET) particles and pozzolanic materials on the properties of self-compacting concrete”, Journal of Cleaner
•
AC C
Production, Vol. 112, pp: 2363-2373. Shahid Iqbal, Ahsan A., Klaus Holschemacher, Thomas A. Bier, (2015), “Mechanical properties of steel fiber reinforced high strength lightweight self-compacting concrete (SHLSCC)”, Journal of Construction and Building Materials, Vol. 98, pp: 325-333. •
Song P.S., Wub J.C., Hwang S., Sheu B.C., (2005a), “Assessment of statistical variations in impact resistance of high-strength concrete and high-strength steel fiberreinforced concrete”, Journal of Concrete and Cement Research, Vol. 35, pp: 393-399.
ACCEPTED MANUSCRIPT •
Song P.S., Wub J.C., Hwang S., Sheu B.C., (2005b),”Statistical analysis of impact strength and strength reliability of steel–polypropylene hybrid fiber-reinforced concrete”, Journal of Construction and Building Materials, Vol. 19, pp: 1-9. Swamy R.N., Stavrides H., (1976), “Some statistical considerations of steel fibre
EP
TE D
M AN U
SC
RI PT
Composites”, Journal of Cement and Concrete Research, Vol. 6, pp: 201-216.
AC C
•
ACCEPTED MANUSCRIPT
(%)
SiO2
21.1
Al2O3
4.37
Fe2O3
3.88
MgO
1.56
K2O
0.52
Na2O
0.39
CaO
63.33
C3S
51
SC
Composition
RI PT
Table 1. Chemical composition and physical properties of Portland cement
C2S
22.7
C3A
5.1
11.9
M AN U
C4AF
Physical properties Specific gravity
3.11
2
AC C
EP
TE D
Specific surface (cm /g)
3000
ACCEPTED MANUSCRIPT
Table 2. Proportions of mix compositions (kg/m3) Water/Cement 0.34 0.34 0.34 0.34
Cement 940 940 940 940
Fiber 0 7 21 35
Silica fume 132 132 132 132
Fine aggregates 940 940 940 940
AC C
EP
TE D
M AN U
SC
RI PT
Specimen designation GFRC0 (Plain self-compacting concrete) GFRC0.25 GFRC0.75 GFRC1.25
SP 3.4 3.4 3.4 3.4
ACCEPTED MANUSCRIPT
Table 3. Compressive strengths of cubic specimens
EP AC C
GFRC1.25 56.12 58.00 54.37 61.47 55.34 67.14 58.75 53.11 48.95 63.95 67.79 56.18 59.58 61.18 52.41 62.86 60.27 61.28 59.64 64.95 59.17 8.38
SC
RI PT
Compressive Strength (MPa) GFRC0.25 GFRC0.75 53.42 60.57 53.74 59.12 52.08 56.52 49.04 54.12 50.16 53.07 52.38 58.56 47.63 53.41 50.37 58.26 51.5 56.54 49.53 54.22 47.69 53.93 52.45 57.1 49.67 54.24 49.35 54.59 45.86 50.75 50.49 56.66 48.59 52.31 47.66 57.79 55.89 62.41 46.75 52.55 50.21 55.83 5.12 5.40
TE D
GFRC0 39.02 41.84 41.52 37.62 --------------------------------40.00 4.10
M AN U
Specimen No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Ave. COV (%)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 4. Flexural strengths of prismatic beams Flexural strength (MPa) Specimen No. GFRC0.25 GFRC0.75 GFRC1.25 1 4.82 5.57 6.8 2 4.63 5.5 5.66 3 5.1 5.96 6.36 4 4.92 5.4 6.57 5 4.05 5.07 5.64 6 5.23 4.93 6.22 7 4.61 4.32 5.94 8 4.57 5.6 5.47 9 5.72 6.01 6.12 10 4.38 5.01 5.56 11 4.5 4.74 7.17 12 4.86 4.76 5.22 13 5.43 6.25 5.31 14 5.03 5.36 5.89 15 4.46 5.66 6.2 16 5.03 5.03 5.93 17 5.46 5.83 6.36 18 4.98 5.2 6.62 19 4.78 5.9 4.84 20 5.15 5.33 5.8 Ave. 4.88 5.36 5.98 COV (%) 8.33 9.24 9.61
ACCEPTED MANUSCRIPT
Table 5. Impact resistance of GFRC specimens (Blow)
EP
AC C
FC 123 131 70 57 111 69 102 48 33 123 66 74 54 72 62 94 43 109 116 141 30 89 34 120 88 24 80 48 131 75 42 99 61 113 51 75 58 51 45 27 75.97 32.98 43.41
GFRC1.25 UC 139 164 93 63 131 87 116 67 46 166 77 96 72 88 78 112 62 148 152 169 43 168 51 174 126 30 94 67 180 122 49 123 74 117 58 94 80 66 53 38 98.32 43.21 43.94
RI PT
INPB 10 13 19 6 11 10 18 10 3 29 5 19 8 13 9 7 6 13 21 11 18 28 19 9 13 19 26 4 9 9 18 19 27 18 29 26 14 14 19 24 15.07 7.34 48.73
SC
GFRC0.75 UC 71 37 62 47 67 34 118 58 45 99 41 87 40 83 66 23 69 78 97 76 60 107 64 47 56 70 120 48 82 36 76 73 123 66 88 78 50 50 103 147 71.05 27.59 38.84
M AN U
FC 61 24 43 41 56 24 100 48 42 70 36 68 32 70 57 16 63 65 76 65 42 79 45 38 43 51 94 44 73 27 58 54 96 48 59 52 36 36 84 123 55.97 22.78 40.69
TE D
Number GFRC0.25 No. FC* UC** INPB 1 37 42 5 2 39 46 7 3 21 30 9 4 28 31 3 5 41 45 4 6 37 42 5 7 36 37 1 8 51 55 4 9 23 26 3 10 50 72 22 11 25 40 15 12 42 45 3 13 30 44 14 14 20 25 5 15 31 38 7 16 43 53 10 17 45 61 16 18 19 23 4 19 42 47 5 20 17 24 7 21 44 62 18 22 51 65 14 23 35 45 10 24 56 60 4 25 25 32 7 26 37 42 5 27 18 26 8 28 32 34 2 29 41 52 11 30 34 36 2 31 46 58 12 32 69 82 13 33 23 29 6 34 33 45 12 35 60 85 25 36 58 83 25 37 74 78 4 38 27 33 6 39 27 32 5 40 56 82 26 Mean 38.07 47.07 9.10 SD 14.00 17.95 6.70 COV (%) 36.77 38.14 73.64 *FC: First crack impact resistance **UC: Ultimate crack impact resistance
INPB 16 33 23 6 20 18 14 19 13 43 11 22 18 16 16 18 19 39 36 28 13 79 17 54 38 6 14 19 49 47 7 24 13 4 7 19 22 15 8 11 22.35 15.46 69.20
ACCEPTED MANUSCRIPT
Table 6. Some statistical results obtained from implemented tests on the specimens Max.
Range***
51.41 57.24 61.48
45.86 50.75 48.95
55.89 62.41 67.79
10.03 11.66 18.84
0.999 0.646 0.999
0.879 0.507 0.978
5.07 5.60 6.25
4.05 4.32 4.84
5.72 6.25 7.17
1.67 1.93 2.33
0.999 0.999 0.999
0.983 0.965 0.997
17 16 24
74 123 141
57 107 117
0.999 0.999 0.685
0.438 0.27 0.096
23 23 30
85 147 180
62 124 150
0.271 0.834 0.568
0.062 0.260 0.053
42.55 63.26 86.52 52.91 79.87 112.14
AC C
EP
KS*: Kolmogorov–Smirnov AD**: Anderson–Darling Range***: Difference between the minimum and maximum
SC
RI PT
Min.
TE D
The compressive strength GFRC0.25 49.00 GFRC0.75 54.42 GFRC1.25 56.84 The flexural strength GFRC0.25 4.69 GFRC0.75 5.14 GFRC1.25 5.71 The first crack impact resistance GFRC0.25 33.59 GFRC0.75 48.68 GFRC1.25 65.42 The ultimate crack impact resistance GFRC0.25 41.44 GFRC0.75 62.22 GFRC1.25 84.50
p-value AD**
Upper 0.95% CI of mean
M AN U
Lower 0.95% CI of mean
KS*
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 7. Number of replications required to retain the error under a specified limit at 90% and 95% level of confidence GFRC0.25 GFRC0.75 GFRC1.25 Error (%) FC UC FC UC FC UC 90% level of confidence <10 23 24 25 28 31 32 <15 10 11 12 13 14 15 <20 6 6 7 7 8 8 <25 4 4 4 5 5 6 <30 3 3 3 4 4 4 <35 2 2 3 3 3 3 <40 2 2 2 2 2 2 <45 2 2 2 2 2 2 <50 1 1 1 2 2 2 95% level of confidence <10 37 40 41 45 51 53 <15 17 18 19 20 23 24 <20 10 10 11 12 13 14 <25 6 7 7 8 9 9 <30 5 5 5 5 6 6 <35 3 4 4 4 5 5 <40 3 3 3 3 4 4 <45 2 2 3 3 3 3 <50 2 2 2 2 3 3
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
Fig 1. Distribution of used fine aggregate
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
Fig 2. Used waste GFRP sheets as reinforcement of self-compacting concrete
ACCEPTED MANUSCRIPT
SC
RI PT
Distributed recycled GFRP fibers
AC C
EP
TE D
M AN U
a) b) Fig 3. a) Slump flow test for reinforced self-compacting concrete; b) Microscopic images to indicate the homogeneity of fibers in the reinforced self-compacting concrete with 1.25% recycled GFRP fibers
EP
TE D
M AN U
a)
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
b) Fig 4. a) Some prismatic beams and the adopted test setup for execution of FPB test; b) Disks and apparatus of impact test
ACCEPTED MANUSCRIPT
T500=1.6949Vf +4.7966 2
M AN U
a)
SC
RI PT
R =0.963
TE D
Tv=2.1017Vf+10.068 2 R =0.857
b)
AC C
EP
Fig 5. Effect of different glass fiber percentage contents on: a) Slump flow and T500; b) TV
SC
RI PT
ACCEPTED MANUSCRIPT
b)
d)
AC C
EP
c)
TE D
M AN U
a)
e) Fig 6. Effect of different glass fiber percentage contents on: a) Compressive strength; b) Flexural strength; c) The first crack impact resistance; d) The ultimate crack impact resistance; e) INPB
GFRC0.75
GFRC1.25
M AN U
SC
GFRC0.25
RI PT
ACCEPTED MANUSCRIPT
TE D
Distributed recycled GFRP fibers
AC C
.
EP
Fig 7. Failed disks under drop weight impact loading
ACCEPTED MANUSCRIPT
Rupture of glass fibers
TE D
M AN U
SC
Cement hydrated products
RI PT
Cement hydrated products
AC C
EP
Fig 8. SEM images from predominated failure modes of recycled glass fiber
Rupture of glass fibers
SC
RI PT
ACCEPTED MANUSCRIPT
ffr (MPa)
b) Flexural strength
FC (Blow) c)
TE D
M AN U
a) Compressive strength
The first crack impact resistance
UC (Blow) d) The ultimate crack impact resistance
AC C
EP
Fig 9. The histogram of the impact resistance and mechanical properties of specimens GFRC0.25
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ffr (MPa)
b) Flexural strength
EP
TE D
a) Compressive strength
c)
AC C
FC (Blow)
The first crack impact resistance
UC (Blow) d) The ultimate crack impact resistance
Fig 10. Probability distribution of the impact resistance and mechanical properties of specimens GFRC0.25
ACCEPTED MANUSCRIPT
ffr=0.1561fc-2.9533 R2=0.974
M AN U
a)
SC
ffr=0.1599fc-3.5576 R2=0.949
RI PT
ffr=0.1151fc-0.8257 R2=0.985
INPB=0.8616fc-38.825 R2=0.941
TE D
INPB=0.9451fc-43.82 R2=0.881
EP
INPB=0.5587fc-23.902 R2=0.923
AC C
b) Fig 11. a) Compressive strength vs. flexural strength; b) Compressive strength vs. INPB
UC=2.5231fc-93.39 R2=0.958
M AN U
a)
SC
FC=2.3901fc-92.96 R2=0.958
RI PT
ACCEPTED MANUSCRIPT
UC=3.9359fc-169.91 R2=0.916
TE D
FC=2.9957fc-128.87 R2=0.869
AC C
EP
b)
UC=2.767fc-115.07 R2=0.974
FC=3.2559fc-130.19 R2=0.987
c) Fig 12. Compressive strength vs. the first or ultimate crack impact resistance for: a) GFRC0.25; b) GFRC0.75; c) GFRC1.25
M AN U
a)
SC
UC=1.258×FC-0.7225 R2=0.965
RI PT
ACCEPTED MANUSCRIPT
TE D
UC=1.205×FC+3.5999 R2=0.989
AC C
EP
b)
UC=1.3043×FC-0.7664 R2=0.991
c) Fig 13. The first crack impact resistance vs. the ultimate crack impact resistance
ACCEPTED MANUSCRIPT Highlights
The impact resistance and mechanical properties of reinforced SCC with rGFRP. The compressive and flexural strength of reinforced SCC follow normal distribution.
AC C
EP
TE D
M AN U
SC
RI PT
The impact resistance and mechanical properties were correlated linearly.
S0959-6526(16)30069-5
DOI:
10.1016/j.jclepro.2016.02.148
Reference:
JCLP 6858
To appear in:
Journal of Cleaner Production
Received Date: 4 November 2015 Revised Date:
23 February 2016
Accepted Date: 26 February 2016
Please cite this article as: Mastali M, Dalvand A, Sattarifard A, The impact resistance and mechanical properties of reinforced self-compacting concrete with recycled glass fiber reinforced polymers, Journal of Cleaner Production (2016), doi: 10.1016/j.jclepro.2016.02.148. 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, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
The impact resistance and mechanical properties of reinforced selfcompacting concrete with recycled glass fiber reinforced polymers
RI PT
M. Mastali1, A. Dalvand2*, A. Sattarifard 3 1. ISISE, Minho University, Department of Civil Engineering, Campus de Azurem, Guimaraes, 4800-058 Portugal
SC
2. Assistant Professor, Department of Engineering, Lorestan University, Khorramabad, Iran
M AN U
Corresponding author Email: [email protected]
3. Faculty of Civil Engineering, Semnan University, Semnan, Iran
Abstract
TE D
Experimental and statistical analyses are represented considering the impact resistance and mechanical properties of self-compacting concrete reinforced with recycled Glass Fiber Reinforced Polymers (GFRP). Specimens were reinforced with recycled GFRP in three
EP
groups, including 0.25%, 0.75%, and 1.25% of fiber volume fractions. An extensive experimental program including two hundred and fifty two specimens were prepared and
AC C
tested to characterize the impact resistance and mechanical properties of the reinforced selfcompacting concrete with glass fiber reinforced polymers. Scanning Electron Microscope images were used to perceive the failure mechanism of recycled glass fiber in the matrix. Then, the relatively large and reliable collected data from the experimental tests was used to start statistical and analytical analysis. The results showed that adding recycled glass fiber reinforced polymers results in improving the impact resistance and mechanical properties of the reinforced self-compacting concrete with glass fiber reinforced polymers. Moreover, the statistical data analysis revealed that the
ACCEPTED MANUSCRIPT impact resistance and mechanical properties of reinforced self-compacting concrete follow a normal distribution. Furthermore, some equations were developed to correlate the impact resistance of the reinforced self-compacting specimens to the mechanical properties with high values of coefficient of determination.
RI PT
Keywords: Recycled GFRP fiber; Self-compacting concrete; Mechanical properties; Impact
AC C
EP
TE D
M AN U
SC
resistance.
ACCEPTED MANUSCRIPT
1. Introduction Using some recyclable waste materials in different products is favourable because it is economic and environmental-friendly and therefore, researchers have started to use recycled materials in the concrete mixes, which is one of the main construction materials.
RI PT
Significant studies have been conducted to use waste materials in concrete using different ways such as reinforcement of concrete with recycled fibers or using waste materials as powder to replace Portland cement (Jianzhuang X. et al., 2015; Abdollahnejad Z. et al.,
SC
2015; Ghasemi N.M. et al., 2014; Foti D., 2013)
Foti D. in 2011 worked on the reinforced concrete with Polyethylene Terephthalate (PET)
M AN U
bottles waste fibers. She found that adding small amounts of recycled fibers from PET bottle wastes could have significant effects on improvement of post-cracking performance of compositions. Moreover, it was revealed that these fibers enhanced the toughness of specimens. In 2013, Foti D. also studied the feasibility of recycling PET fibers, obtained from
could increase ductility.
TE D
waste bottles with different shapes. Her studies revealed that using PET fibers in concrete
The applicability of self-compacting concrete containing waste Polyethylene Terephthalate (PET) particles was assessed using slump flow, V-funnel and L-box tests. Moreover,
EP
mechanical properties of mixtures including compressive, tensile and flexural strengths, and
AC C
modulus of elasticity were evaluated. The experimental results indicated that waste PET particles could be reused as aggregates in self-compacting concrete, while using the waste PET particles led to degrading mechanical properties. However, fly ash and silica fume compensated the loss of strength caused by adding PET (Sadrmomtazi A. et al., 2016). Using Fiber Reinforced Polymer (FRP) materials has increased significantly in several applications, including industry, construction, and transportation due to its light weight, high stiffness, high tensile strength, good durability resistance, and high temperature resistance. In this regard, many products are recently manufactured by FRP materials in the construction industry such as sheets for strengthening structural elements (like beams,
ACCEPTED MANUSCRIPT columns, and slabs), bars to reinforce concrete due to good properties of FRP materials against corrosion, structural profiles, and sandwich panels (Mastali M. et al., 2015). Production process usually results in some waste FRP materials, which should be recycled through material recycling, chemical recycling, or thermal recycling. Otherwise, they are
RI PT
presumed to be buried in the ground or remain unusable. Asokan P. et al. investigated the recycling potential of glass fibre reinforced plastic waste in concrete and cement composites. Results revealed that the compressive strength of specimens using 5%–50% Glass Fibre Reinforced Plastic (GFRP) waste powder under
SC
water curing varied from 19 MPa to 37 MPa. Increasing the content of GRP waste decreased the compressive strength. The density of concrete with 50% GRP waste reduced
M AN U
about 12% compared to the reference concrete. Moreover, the flexural strength for specimens using 5% GRP waste fibre was recorded as 16.5 MPa, while the flexural strength for control beam was measured as 10.5 MPa.
Keilji et al. in 2005 studied mechanical properties of concrete reinforced with Carbon Fiber
TE D
Reinforced Polymer (CFRP) fibers. The applied CFRP fibers were provided from waste CFRP sheets. In their study, they investigated the effects of size and content of chopped CFRP fibers on the compressive and flexural strengths. They indicated that the improvement
EP
of mechanical properties of reinforced concrete with CFRP fibers is governed significantly by the size of fibers. In 2014, García D. et al. worked on evaluation of mechanical properties of
AC C
recycled Glass Fiber Reinforced Plastics (rGFRP) from different sources as short fibers. The physical–mechanical properties of the concrete reinforced with rGFRP were characterized by conducting compressive and flexural tests as well as shrinkage stability and alkaliaggregate reaction. Moreover, the interaction between the rGFRP and the matrix was observed and characterized by Scanning Electron Microscope (SEM) images. They found that the compressive and flexural strength increased 22% and 16% respectively compared to control specimens, after providing rGFRP materials from an optimized milling process (García D. et al., 2014).
ACCEPTED MANUSCRIPT Successful application of recycled fibers to reinforce plain concrete enhances the impact resistance, and improves mechanical and durability properties. However, this improvement strongly depends on the composition, and also mechanical properties and contents of recycled fibers. Therefore, some studies have been performed on mechanical properties of
RI PT
reinforced concrete with some types and contents of recycled fibers under static loading. Since the impact resistance and mechanical properties of reinforced self-compacting concrete with recycled fibers are not characterized completely yet, in the present paper, an experimental/statistical study was conducted to characterize the impact resistance and
SC
mechanical properties of self-compacting concrete reinforced with recycled glass fiber in different volume fractions, including 0.25%, 0.75%, and 1.25%. Considering statistical
M AN U
analysis is appropriate for better understanding the impact resistance and mechanical properties of reinforced self-compacting concrete. To the best of the author’s knowledge, an extensive experimental/statistical study on impact resistance and mechanical properties of reinforced self-compacting concrete made with rGFRP fibers has not been reported yet.
TE D
Although some studies have considered mechanical properties of reinforced normal concrete with recycled GFRP fibers (García et al., 2014). Yet, they have not included a statistical assessment based on a relatively large number of specimens under compression,
AC C
EP
flexure, and impact test.
2. Experimental program 2.1.
Materials and concrete mix design
In this study, mixture compositions are a combination of Portland cement 42.5R, fine aggregates, silica fume, water, and superplasticizer (SP). Table 1 represents the chemical and physical properties of Portland cement 42,5R. Also, the proportions of materials used in the mixtures are mentioned in Table 2. The contents of materials in Table 2 have been obtained according to a target slump flow equal or more than 600 mm for plain self-
ACCEPTED MANUSCRIPT compacting concrete. In this regard, some preliminary experimental tests were implemented on workability of some plain self-compacting concretes. The results revealed that using cement to fine aggregate ratios equal to one resulted in slump flows of lower than 600 mm.
Table 1.
RI PT
Table 2. The particle distribution of fine aggregate is shown in Fig. 1. The workability of mixtures was adjusted through a high range water reducer agent with the commercial name of Mape110.
SC
Moreover, GFRP fibers were recycled from the remained unusable GFRP sheets in the Civil Lab of Semnan University. In this regards, GFRP sheets were cut to smaller GFRP fibers
M AN U
with an average length of 20 mm using a shred machine, as shown in Fig. 2. The thickness of GFRP sheets was measured as 0.11 mm. The tensile strength, elastic modulus, and density of the GFRP sheets, used to produce the GFRP fibers, were 1103 MPa, 44.8 GPa, and 2080 kg/m3, respectively. All efforts were made to provide equal recycled glass fiber lengths.
TE D
Fig 1. Fig 2.
To batch, cement was mixed with fine aggregate for one minute. Then, the water and super
EP
plasticizer were added to the composition that was mixed for 6–8 min. Finally, the concrete-
AC C
fibers mixtures were prepared by inchmeal addition of the glass fibers to the fresh selfcompacting concrete while adding enough fibers to concrete mixture so that the desired ratio of fibers to volume was achieved. Chopped glass fibers were incrementally added to the self-compacting concrete to avoid unfavourable effects such as balling fibers. Moreover, the homogeneity of the composition was assessed by analysing the microstructure of mixtures through microscopic images. This analysis was executed on the hardened reinforced concrete. Then, mixed self-compacting concrete was cast into cubic molds (100×100×100 mm), cylindrical disk molds (150×65 mm), and prismatic beams (500×60×60 mm) to be tested in compressive, impact, and flexural tests, respectively. The specimens were sealed
ACCEPTED MANUSCRIPT and retained at 250C and 90% relative humidity for 24 hours. After demolding, specimens were cured in water at 230C for 28 days and all specimens were tested afterwards. The fresh state behavior of plain self-compacting concrete was assessed using the slump test and the V funnel test (EN 12350-8:2010, 2010; EN 12350-9: 2010, 2010; Shahid Iqbal et
Fig 3.
Test setups and test procedures
2.2.1.
Compressive test
SC
2.2.
RI PT
al., 2015). Fig. 3a presents the slump flow test for reinforced self-compacting concrete.
According to the compositions listed in Table 2, sixty-four cubic specimens with dimensions
M AN U
of 100 ×100 × 100 mm were cast and tested in three different reinforced groups to study the effects of different recycled glass fibers volumes on the compressive strength, including 0.25%, 0.75%, and 1.25% recycled glass fibers. Four cubic specimens were used to evaluate the compressive strength of plain self-compacting concrete and twenty cubic specimens were employed to assess the compressive strength of each group of reinforced
2.2.2. Flexural test
TE D
specimens with rGFRP fibers.
EP
To assess the flexural performance of reinforced self-compacting concrete with recycled
AC C
glass fibers, sixty-four prismatic beams (60 × 60 × 500 mm) were cast and tested under Four Point Bending (FPB) test. The test setup used to perform flexural assessment is depicted in Fig. 4a.
Fig 4.
The flexural loading was applied to the beams with a loading rate of 0.5 mm/min. Moreover, a Linear Variable Differential Transformer (LVDT) of 10 mm stroke was used to record midspan deflection. As mentioned before, fewer variables are involved in mechanical properties of plain self-compacting concrete compared to reinforced self-compacting concrete. In this regard, four prismatic specimens were used to assess the flexural strength of plain self-
ACCEPTED MANUSCRIPT compacting concrete, while twenty prismatic specimens were required for each group of prismatic beams reinforced with rGFRP fibers.
2.2.3. Impact drop weight test
RI PT
The drop weight impact test was conducted based on ACI Committee 544 on fiber reinforced concrete. The apparatus is depicted in Fig. 4b (ACI Committee 544, 1988), in which a steel hammer of 4.45 kg drops from a height of 457 mm on a steel ball with a diameter of 63.5
SC
mm. This steel ball is in contact with central surface of specimens. Totally, one hundred and twenty four disks with a diameter of 150 mm and a height of 65 mm were cast and prepared
M AN U
to test at the age of 28 days. To calculate energy absorption, the following equation is used: Impact energy (En)= N × W × H
(1)
where, N is number of blows, W is weight of steel hammer with a mass of 4.45 kg, and H is the height of fall. To assess the impact resistance of disks under drop weight impact load, four cylinders were used to evaluate the impact resistance of plain concrete, while for each
TE D
group of reinforced cylinder with rGFRP fibers, forty cylinders were used.
3.1.
EP
3. Results and discussion Fresh mix properties
AC C
The homogeneity of mixtures was investigated by analysing the microstructure of the prepared mixes. Fig. 3b shows that the rGFRP fibers are distributed homogeneously regardless of the fiber content, considerable amounts of the rGFRP fibers well attached to the cement-based matrix.
Fig. 5 presents the results recorded for the fresh state properties of concrete mixtures. Fig. 5a shows that increasing rGFRP fiber results in reduction of the slump flow, so that the maximum reduction in slump flow is recorded as 15% for specimen GFRC1.25. Moreover, it was found that Tv and T500 increased linearly as the fiber content increased.
ACCEPTED MANUSCRIPT Fig 5. Preliminary experimental tests were also carried out on the workability of the reinforced selfcompacting concrete with 2% recycled GFRP fiber. The obtained results for the reinforced self-compacting concrete with 2% recycled GFRP fiber showed that the workability reduced
RI PT
significantly compared to the reinforced self-compacting concrete with 1.25% rGFRP fiber. Therefore, the study focused on the reinforced self-compacting concrete specimens with
SC
GFRP fiber content of lower than 2%.
3.2.1. Compressive strength
M AN U
3.2. Hardened concrete properties
As previously mentioned, the evaluation of compressive strength of cubic specimens was carried out using sixty-four cubic specimens of 100 × 100 × 100 mm. A digital standard
TE D
automatic testing machine with a capacity of 600 kN was used to apply compressive loading. The compressive strength of specimens was measured and recorded in Table 3. Table 3.
EP
The compressive strength of specimens shows that reinforcement of plain self-compacting
AC C
concrete results in improving the compressive strength, so that this improvement is significantly governed by recycled glass fiber contents. Adding glass fiber leads to bridging action in the reinforced cement-based composites. Transferring stress from glass fibers to the matrix regard to interfacial shear strength offers more resistance to crack opening in the specimens. As shown in Fig. 6a, reinforcement of plain self-compacting concrete resulted in an increase of 25.52%, 39.57%, and 47.92% in compressive strength of specimens with recycled glass fiber of 0.25%, 0.75%, and 1.25%, respectively. This improvement could be attributed to arresting cracks in specimens by glass fibers under compressive loading. Fig 6.
ACCEPTED MANUSCRIPT Fig 9a represents the histogram of compressive strength of specimen GFRC0.25. The indicated results depicts that the compressive strengths of specimens are normally distributed and fit well with the superimposed normal distribution curve (more discussions on statistical analysis will be presented in section 4.2). GFRC1.25 had the highest mean
RI PT
compressive strength (59.17 MPa) among all the groups, while the highest coefficient of variation (COV) of 8.38% belonged to this group. Additionally, as it could be seen in Table 3, increasing glass fiber content results in increased scatter of compressive strength.
SC
3.2.2. Flexural strength
M AN U
Sixty-four prismatic beams of 500 × 60 × 60 mm were employed to be assessed under Four Point Bending (FPB) test. Table 4 indicates the registered flexural strengths of prismatic beams. Considering the recorded results, increasing glass fiber content leads to enhanced flexural performance and increased flexural strength of beams. Fig. 6b presents the efficiency of recycled glass fiber on increasing the flexural strength of specimens.
TE D
Reinforcement of plain self-compacting concrete beams with recycled fiber volume fractions of 0.25%, 0.75%, and 1.25% led to 30.13%, 42.93%, and 59.46% increase in the flexural strength of specimens, respectively (see Fig. 6b). Additionally, the maximum flexural
EP
strength was measured as 5.98 MPa for specimen GRPC1.25, which is 22.54% and 11.56% more than mean flexural strengths of specimens with 0.25% and 0.75% recycled GFRP
AC C
fibers, respectively. Moreover, it was observed that like compressive strength, increasing the content of recycled glass fiber leads to increased scatter of the results obtained for flexural strengths, so that the maximum COV was recorded as 9.61% for specimens reinforced with 1.25% of glass fiber.
Table 4. Enhancement of flexural performance of reinforced beams with recycled GFRP fibers can be due to the bridging action of GFRP fiber, through which the bridging fibers can partially transfer the stress across the crack. In 2015, Felekoglu et al. found that increasing
ACCEPTED MANUSCRIPT homogeneous distribution of fibers throughout the composite may also increase the number of formed cracks. The frequency histogram of flexural strength of specimen GFRC0.25 is indicated in Fig 9b. This figure shows that flexural strengths follow a normal distribution.
RI PT
3.2.3. Impact resistance The drop weight impact test was conducted on one hundred twenty four cylinder disks. The recorded impact results are presented in Table 5. Reinforcement of plain self-compacting
SC
concrete with recycled glass fiber of 0.25%, 0.75%, and 1.25% increased the first crack impact resistance of disks 2.53, 3.73, and 5.06 times, respectively. Furthermore, as shown in
M AN U
Figs 6c and 6d, the ultimate crack impact resistance increased 2.94, 4.44, 6.14 times for reinforced cylinders with recycled glass fiber volume fractions of 0.25%, 0.75%, and 1.25%, respectively. Like the results attained for mechanical properties of reinforced specimens, reinforcement of disks resulted in increasing the impact resistance, so that both the first crack impact resistance and the ultimate crack impact resistance increased significantly.
TE D
Table 5.
As listed in Table 5, the maximum value of the first crack impact resistance and the ultimate impact resistance were recorded 75.97 and 98.32 blows for reinforced disks with 1.25%
EP
glass fiber, respectively.
By comparing the recorded first and ultimate crack impact resistance of specimens, it is
AC C
revealed that the first and ultimate crack impact resistance of disks reinforced with 1.25% of glass fiber increase 35.73% and 38.38% as compared to the first and ultimate crack impact resistance of reinforced disks with 0.75% of recycled glass fiber, respectively. Moreover, this increase is 99.55% and 2.08 times for specimens reinforced with 0.25% fiber. As it occurred for mechanical properties of specimens, increasing glass fiber content results in increased scatter of the first and ultimate impact resistance, so that the maximum COV for the first and ultimate crack resistance were recorded as 43.41% and 43.49% for specimens of GFRC1.25, respectively.
ACCEPTED MANUSCRIPT ACI Committee 544’s repeated drop-weight impact test for concrete is often criticized for large variations within the results, therefore, a modified impact test was proposed to ACI test thatsignificantly improved the reliability of the results (Badr et al., 2005). The impact test in the present paper was implemented based on ACI Committee 544, therefore, the recorded
RI PT
large scatter can be derived from large variations involved in the repeated drop-weight impact test.
Table 6 presents some statistical results from experimental data. It shows that increasing fiber content results in increasing the interval difference between the minimum and
resistance of specimens.
M AN U
Table 6.
SC
maximum values of compressive and flexural strengths, and also the first/ultimate impact
The number of blows to reach failure impact resistance of disks varied from 38.07 to 47.07 for the specimens GFRC0.25, from 55.97 to 71.05 for the specimens of GFRC0.75, and from 75.97 to 98.32 for the specimens of GFRC1.25. Using more recycled glass fiber increased
TE D
the difference between blow number of the first crack impact resistance and the blow number of the ultimate crack impact resistance. Here in this study, increasing the Number of Post initial crack Blows to failure is called ‘‘INPB’’ parameter. Increasing glass fiber content
EP
from 0.25 to 0.75% and 1.25% increased the number of post-initial crack blows to failure 67.55% and 2.48 times, respectively. Moreover, reinforcement of plain self-compacting
AC C
concrete with 0.25%, 0.75%, and 1.25% increased the number of post-initial crack blows to failure of 9, 15.08, and 22.35 times, respectively. Based on equation (1), the maximum absorbed energy for the first and ultimate impact resistances were calculated as 1518.92 J and 1965.00 J for specimens of GFRC1.25, respectively. The frequency histograms of the first and ultimate impact resistance of specimen GFRC0.25 are shown in Figs 9c and 9d, which indicate that the first and ultimate impact resistance of specimens hardly follow a normal distribution due to low p-values listed in Table 6. Fig. 7 depicts the formed crack patterns on some tested disks, which were reinforced with 0.25%, 0.75%, and 1.25% recycled glass fiber. Increasing fiber content resulted in more cracks on the surface of
ACCEPTED MANUSCRIPT specimens, as shown in Fig. 7. Forming multiple cracks is due to fiber bridging action of glass fibers. Additionally, a fractured surface is indicated in Fig. 7, where parts of the recycled glass fibers are marked. Respect to the presented fractured surface, it is revealed that recycled glass fibers are distributed uniformly in the section.
RI PT
Fig 7. SEM images were used to reach a profound understanding about the fiber reinforcement mechanisms. As the desired bonding was formed between glass fibers to cement based
SC
materials, cement-hydrated particles almost covered the surface of fibers, as shown in Fig. 8. The SEM images obtained in the fracture surface of tested specimens show a tendency
M AN U
for the rupture of the long glass fibers. Fig 8.
4.
Statistical and analytical analysis
TE D
In this section, probable distribution of compressive strength, flexural strength, and impact resistance of recycled glass fiber reinforced self-compacting concretes is discussed. Most statistical data analyses in the literature are related to mechanical properties of reinforced
EP
concrete made with the steel and Polypropylene (PP) fibers (Song P.S. et al., 2005a; Song P.S. et al. 2005b; Natarajaa M.C. et al., 1999; Fakharifar M. et al., 2014; Rahmani T. et al.,
AC C
2012) and no study has been conducted on mechanical properties of self-compacting concrete reinforced with recycled GFRP fibers so far. Analysing a large experimental database provided a reliable source for implementing statistical studies. Statistical analysis establishes a better understanding of mechanical properties of these materials. Effect of type and volume fraction of fiber on statistical parameters and distribution of flexural strength, compressive strength, and impact strength are important to be described. Moreover, some empirical relations are developed for mechanical properties of specimens, as well as compressive strength of specimens used to predict impact resistance of specimens.
ACCEPTED MANUSCRIPT
4.2.
Probability distribution
To conduct the statistical analysis, the OriginLab computer program was used to find the probability distributions of mechanical properties. To test the normality of data, Kolmogorov–
RI PT
Smirnov (KS test) and Anderson–Darling (AD test) methods were used. Kolmogorov– Smirnov test concentrates on the maximum deviation between the observed cumulative histogram and the hypothesized cumulative distribution function (Jack R. Benjamin et al.,
SC
1970). This feature of fit test evaluates the feasibility of using the normal distribution for the impact resistance and mechanical properties of fiber reinforced self-compacting concrete.
M AN U
For both methods, a p-value is computed to quantify the strength of the evidence against the null hypothesis (Rahmani T. et al., 2012). A p-value of smaller than 0.05 indicates that the null hypothesis (confirming normal distribution) is rejected. The hypotheses for the normal distribution are as follows:
Ho1. The compressive strength of recycled glass fibre reinforced self-compacting concrete
TE D
follows the normal distribution;
Ho2. The flexural strength of recycled glass fibre reinforced self-compacting concrete follows the normal distribution;
EP
Ho3. The first crack impact resistance of recycled glass fibre reinforced self-compacting concrete follows the normal distribution;
AC C
Ho4. The ultimate crack impact resistance of recycled glass fibre reinforced self-compacting concrete follows the normal distribution; Ha. Not Ho1, 2, 3, 4.
The obtained p-values for all tests are listed in Table 6, which are all above 0.05. This finding corroborates the null hypothesis at the 0.05 significance level. Therefore, the compressive strength, flexural strength, first crack impact resistance, and ultimate crack impact resistance of reinforced SCC with rGFRP follow the normal distribution. It is worth stating that mechanical properties of the specimens are better fitted to the normal distribution compared
ACCEPTED MANUSCRIPT to the first and the ultimate crack impact resistance of the specimens, due to higher p-values of the compressive strength and flexural strength compared to the p-values registered for the first crack impact resistance, ultimate impact resistance. This result confirms the findings of Fakharifar et al. (Fakharifar M. et al., 2014; Rahmani T. et al., 2012). Fig. 10 indicates the
RI PT
normal probability curves of the impact resistance and mechanical properties of GFRC specimens. As shown in Fig. 10, the nearly linear pattern of the data of compressive and flexural strengths implies that mechanical properties of GFRC specimens follow a normal
Fig 9.
M AN U
Fig 10.
SC
distribution.
The coefficients of variation of the specimens, tested under drop weight impact test, are listed in Table 5. Table 5 could be used to determine the minimum replication number of tests ‘‘n’’, required for ensuring that the error percentage ‘‘e’’ in the measured average value is below a certain limit. This number of tests can be calculated by equation (2) (Swamy R.N.
TE D
et al., 1976):
2 [ COV ] t 2 n=
e2
(2)
EP
Where, “t” in equation (2) is the value of Student t-distribution for a certain level of confidence, “e” is the error percentage (10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, and
AC C
50%), and COV is the coefficient of variation. The t-distribution is a probability distribution used to estimate population parameters when the sample size is small and/or when the population variance is unknown. The value of Student t-distribution depends on the level of confidence and the degree of freedom, which is related to the number of samples. The value of “t” for degrees of freedom of more than 120 can be considered as 1.645 and 1.282 at 95% and 90% level of confidence, respectively (Box GEP et al., 1978; Moore D.S. et al., 1989). Table 7 shows the minimum number of replications required to keep the error on the specified limits at 90% level of confidence. Table 7 also illustrates that if the error is lower than 10% for the first crack impact resistance, the minimum number of required replications
ACCEPTED MANUSCRIPT would be 23, 25, and 31 for specimens of GFRC0.25, GFRC0.75, and GFRC1.25, respectively at 90% level of confidence. Moreover, for the ultimate crack impact resistance, the minimum number of required replications would be 24, 28, and 32 for specimens of GFRC0.25, GFRC0.75, and GFRC1.25, respectively at 90% level of confidence. This table
RI PT
also indicates if the error remains under 10%, the minimum numbers of replications for GFRC0.25, GFRC0.75, and GFRC1.25 are 37, 41, and 51 for the first-crack strength and 40, 45 and 53 for the failure strength, respectively at 95% level of confidence. According to the
required at each level of error.
4.3.
M AN U
Table 7.
SC
results represented in Table 7, increasing fiber content increases the number of tests
Analytical study
According to the relatively large collected experimental database, the impact resistance and mechanical properties of reinforced concrete self-compacting specimens can be correlated through empirical relations with high coefficient of determination (R2). These empirical
TE D
relations were developed using regression analysis. Fig. 11a indicates the linear empirical equations that relate the flexural strength to the compressive strength. Based on the obtained equations, flexural strengths can be correlated
EP
linearly to compressive strengths for each group of reinforced self-compacting concrete specimens. The slope of developed equation for GFRC0.75 is higher than specimens of
AC C
GFRC0.25 and GFRC1.25. It should be noted that adding 0.75% glass fiber results in the highest rate of flexural strength. Fig 11.
Moreover, based on Fig. 11b, it was revealed that the highest rate of gaining the INPB is achieved for reinforced disks containing 0.75% recycled glass fiber. Fig. 12 depicts the correlation between the compressive strength and the first or ultimate crack impact resistance. The results demonstrate that there is a linear relationship between compressive strength and the first/ultimate crack impact resistance of specimens. Moreover, the ultimate
ACCEPTED MANUSCRIPT impact resistance of GFRC0.75 had the highest slope followed by GFRC1.25 and GFRC0.25. It means that specimens of GFRC0.75 have the best impact performance. Fig 12. Furthermore, the relation between the first impact resistance versus the ultimate crack
RI PT
resistance of specimens is shown in Fig. 13. A linear relation between the first and ultimate crack resistance of disks is depicted in Fig. 13 with coefficient of determination (R2) of more than 0.96. Linear relationships between the first and ultimate crack impact resistance is also reported (Fakharifar M. et al., 2014) for high performance fiber reinforced concrete with
SC
polypropylene (PP) fibers. According to the recorded results, after forming the first crack on
M AN U
the reinforced specimens, the best performance in increasing the ultimate crack impact resistance was detected in GFRC1.25 specimens. Fig 13.
5.
Conclusion
TE D
An experimental and analytical study was established to investigate the effects of adding recycled GFRP fibers on the impact resistance and mechanical properties of self-compacting concrete. Therefore, two hundred and fifty two specimens were tested experimentally under
EP
compressive test, flexural test, and drop weight impact test and the obtained experimental results were interpreted. Moreover, using SEM images, the failure mechanism of recycled
AC C
glass fiber in the matrix was characterized. Moreover, a statistical data analysis was carried out to better understand the impact resistance and mechanical properties of these materials. In this regard, the relatively large experimental database was used for statistical and analytical analysis and consequently the following results were concluded: 1. Increasing GFRP fiber content resulted in linearly reduced slump flow, while T500 and Tv increased linearly.
ACCEPTED MANUSCRIPT 2. Adding GFRP fiber improves mechanical properties and the impact resistance, so that the maximum compressive strength, flexural strength, and impact resistance were recorded for specimen GFRP1.25. 3. The maximum rate of gaining flexural strength was recorded for GFRC0.75 and the
RI PT
best performance in terms of the first crack and the ultimate impact resistance was recorded for specimens GFRC1.25 and GFRC0.75, respectively.
4. The highest rate of gaining the number of post initial crack blows to failure was obtained for GFRC0.75.
SC
5. Flexural strength, INPB, the first and ultimate impact resistance linearly correlated to compressive strength to the coefficient of determination (R2) of more than 0.86.
M AN U
6. The first and ultimate crack impact resistance are correlated linearly. 7. Increasing glass fiber leads to increased scatter of the results for mechanical properties and impact resistance that can affect fiber/matrix interfacial bond. 8. The compressive strength and flexural strength of the specimens follow the normal
TE D
distribution, while the first and the ultimate crack impact resistance hardly follow a
References
Abdollahnejad Z., Pacheco-Torgal F., Félix T., Tahri W., Barroso Aguiar J., (2015), “Mix
AC C
•
EP
normal distribution.
design, properties and cost analysis of fly ash-based geopolymer foam”, Journal of Construction and Building Materials, Vol. 80, pp: 18-30. •
ACI Committee 544, (1988), Measurement of properties of fiber reinforced concrete. ACI 555 Material Journal, Vol. 85, pp: 583–593.
•
Asokan P., Osmani M., Price A.D.F., (2009), “Assessing the recycling potential of glass fibre reinforced plastic waste in concrete and cement composites”, Journal of Cleaner Production, Vol. 17, pp: 821-829.
ACCEPTED MANUSCRIPT •
Badr A., Ashour F. Ashraf, (2005), “Modified ACI Drop-Weight Impact Test for Concrete”, ACI materials Journal, Vol. 102, pp: 249-255.
•
Box GEP, Hunter WG, Hunter J.S., (1978), “Statistics for experimenters”, Wiley, USA.
•
Burak Felekoglu, Kamile Tosun-Felekog, Eren Gödek, 2015, “A novel method for the
RI PT
determination of polymeric micro-fiber distribution of cementitious composites exhibiting multiple cracking behavior under tensile loading”, Journal of Construction and Building Materials, Vol. 86, pp: 85-94.
EN 12350-8:2010. Testing fresh concrete. Part 8: Self-compacting concrete. Slump-
SC
•
flow test. 2010.
EN 12350-9:2010. In: Testing fresh concrete. Part 9: Self-compacting concrete. V-
M AN U
•
funnel test. 2010. •
Fakharifar M., Dalvand A., Arezoumandi M., Sharbatdar M.K., Chen G., Kheyroddin A., (2014), Mechanical properties of high performance fibre reinforced cementitious composites, Journal of Construction and Building Materials, Vol. 71, pp: 510-520. Foti D., (2011), “Preliminary analysis of concrete reinforced with waste bottles PET
TE D
•
fibers”, Journal of Construction and Building Materials, Vol. 25, pp: 1906-1915. •
Foti D., (2013), “Use of recycled waste pet bottles fibers for the reinforcement of
•
EP
concrete”, Journal of Composite Structure, Vol. 96, pp: 396–404. García D., Vegas I., Cacho I., (2014), “Mechanical recycling of GFRP waste as short-
AC C
fiber reinforcements in microconcrete”, Journal of Construction and Building Materials, Vol. 64, pp: 293-300. •
Ghasemi, N.M., Sharbatdar, M.K., Mastali, M., (2014), “Repairing reinforced concrete slabs using composite layers”, Journal of Materials and Design, Vol. 58, pp: 136-144.
•
Jianzhuang X., Long Li, Luming Shen, Chi Sun Poon, (2015), “Compressive behaviour of recycled aggregate concrete under impact loading”, Journal of Cement and Concrete Research, Vol. 71, pp: 46-55.
ACCEPTED MANUSCRIPT •
Jack R Benjamin, C. Allin Cornell, (1970), Probability, Statistics, and Decision for Civil Engineers, Dover publication, New York, USA.
•
Keiji O., Tomoyuki S., Makoto M., (2005), “Strength in concrete reinforced with recycled CFRP pieces”, Journal of Composite Part A: Applied Science and Manufacturing, Vol.
•
RI PT
36, pp: 893-902. Mastali M., Valente I.B., Joaquim A.O. Barros, Delfina M.F. Gonçalves, (2015), “Development of innovative hybrid sandwich panel slabs: Experimental results”, Journal
•
SC
of Composite Structures, Vol. 133, pp: 476-498.
Moore D.S., McCabe G.P., (1989), “Introduction to the practice of statistics”, New York:
•
M AN U
W.H. Freeman & Company, USA.
Natarajaa M.C., Dhangb N., Guptab A.P., (1999), “Statistical variations in impact resistance of steel fiber-reinforced concrete subjected to drop weight test”, Journal of Concrete and Cement Research, Vol. 29, pp: 989-995.
•
Rahmani T., Kiani B., Shekarchi M., Safari A., (2012), Statistical and experimental
TE D
analysis on the behavior of fibre reinforced concretes subjected to drop weight test, Journal of Construction and Building Materials, Vol. 37, pp: 360-369. •
Sadrmomtazia A., Dolati-Milehsaraa S., Lotfi-Omrana O., Sadeghi-Nikb A., (2016),
EP
“The combined effects of waste Polyethylene Terephthalate (PET) particles and pozzolanic materials on the properties of self-compacting concrete”, Journal of Cleaner
•
AC C
Production, Vol. 112, pp: 2363-2373. Shahid Iqbal, Ahsan A., Klaus Holschemacher, Thomas A. Bier, (2015), “Mechanical properties of steel fiber reinforced high strength lightweight self-compacting concrete (SHLSCC)”, Journal of Construction and Building Materials, Vol. 98, pp: 325-333. •
Song P.S., Wub J.C., Hwang S., Sheu B.C., (2005a), “Assessment of statistical variations in impact resistance of high-strength concrete and high-strength steel fiberreinforced concrete”, Journal of Concrete and Cement Research, Vol. 35, pp: 393-399.
ACCEPTED MANUSCRIPT •
Song P.S., Wub J.C., Hwang S., Sheu B.C., (2005b),”Statistical analysis of impact strength and strength reliability of steel–polypropylene hybrid fiber-reinforced concrete”, Journal of Construction and Building Materials, Vol. 19, pp: 1-9. Swamy R.N., Stavrides H., (1976), “Some statistical considerations of steel fibre
EP
TE D
M AN U
SC
RI PT
Composites”, Journal of Cement and Concrete Research, Vol. 6, pp: 201-216.
AC C
•
ACCEPTED MANUSCRIPT
(%)
SiO2
21.1
Al2O3
4.37
Fe2O3
3.88
MgO
1.56
K2O
0.52
Na2O
0.39
CaO
63.33
C3S
51
SC
Composition
RI PT
Table 1. Chemical composition and physical properties of Portland cement
C2S
22.7
C3A
5.1
11.9
M AN U
C4AF
Physical properties Specific gravity
3.11
2
AC C
EP
TE D
Specific surface (cm /g)
3000
ACCEPTED MANUSCRIPT
Table 2. Proportions of mix compositions (kg/m3) Water/Cement 0.34 0.34 0.34 0.34
Cement 940 940 940 940
Fiber 0 7 21 35
Silica fume 132 132 132 132
Fine aggregates 940 940 940 940
AC C
EP
TE D
M AN U
SC
RI PT
Specimen designation GFRC0 (Plain self-compacting concrete) GFRC0.25 GFRC0.75 GFRC1.25
SP 3.4 3.4 3.4 3.4
ACCEPTED MANUSCRIPT
Table 3. Compressive strengths of cubic specimens
EP AC C
GFRC1.25 56.12 58.00 54.37 61.47 55.34 67.14 58.75 53.11 48.95 63.95 67.79 56.18 59.58 61.18 52.41 62.86 60.27 61.28 59.64 64.95 59.17 8.38
SC
RI PT
Compressive Strength (MPa) GFRC0.25 GFRC0.75 53.42 60.57 53.74 59.12 52.08 56.52 49.04 54.12 50.16 53.07 52.38 58.56 47.63 53.41 50.37 58.26 51.5 56.54 49.53 54.22 47.69 53.93 52.45 57.1 49.67 54.24 49.35 54.59 45.86 50.75 50.49 56.66 48.59 52.31 47.66 57.79 55.89 62.41 46.75 52.55 50.21 55.83 5.12 5.40
TE D
GFRC0 39.02 41.84 41.52 37.62 --------------------------------40.00 4.10
M AN U
Specimen No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Ave. COV (%)
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 4. Flexural strengths of prismatic beams Flexural strength (MPa) Specimen No. GFRC0.25 GFRC0.75 GFRC1.25 1 4.82 5.57 6.8 2 4.63 5.5 5.66 3 5.1 5.96 6.36 4 4.92 5.4 6.57 5 4.05 5.07 5.64 6 5.23 4.93 6.22 7 4.61 4.32 5.94 8 4.57 5.6 5.47 9 5.72 6.01 6.12 10 4.38 5.01 5.56 11 4.5 4.74 7.17 12 4.86 4.76 5.22 13 5.43 6.25 5.31 14 5.03 5.36 5.89 15 4.46 5.66 6.2 16 5.03 5.03 5.93 17 5.46 5.83 6.36 18 4.98 5.2 6.62 19 4.78 5.9 4.84 20 5.15 5.33 5.8 Ave. 4.88 5.36 5.98 COV (%) 8.33 9.24 9.61
ACCEPTED MANUSCRIPT
Table 5. Impact resistance of GFRC specimens (Blow)
EP
AC C
FC 123 131 70 57 111 69 102 48 33 123 66 74 54 72 62 94 43 109 116 141 30 89 34 120 88 24 80 48 131 75 42 99 61 113 51 75 58 51 45 27 75.97 32.98 43.41
GFRC1.25 UC 139 164 93 63 131 87 116 67 46 166 77 96 72 88 78 112 62 148 152 169 43 168 51 174 126 30 94 67 180 122 49 123 74 117 58 94 80 66 53 38 98.32 43.21 43.94
RI PT
INPB 10 13 19 6 11 10 18 10 3 29 5 19 8 13 9 7 6 13 21 11 18 28 19 9 13 19 26 4 9 9 18 19 27 18 29 26 14 14 19 24 15.07 7.34 48.73
SC
GFRC0.75 UC 71 37 62 47 67 34 118 58 45 99 41 87 40 83 66 23 69 78 97 76 60 107 64 47 56 70 120 48 82 36 76 73 123 66 88 78 50 50 103 147 71.05 27.59 38.84
M AN U
FC 61 24 43 41 56 24 100 48 42 70 36 68 32 70 57 16 63 65 76 65 42 79 45 38 43 51 94 44 73 27 58 54 96 48 59 52 36 36 84 123 55.97 22.78 40.69
TE D
Number GFRC0.25 No. FC* UC** INPB 1 37 42 5 2 39 46 7 3 21 30 9 4 28 31 3 5 41 45 4 6 37 42 5 7 36 37 1 8 51 55 4 9 23 26 3 10 50 72 22 11 25 40 15 12 42 45 3 13 30 44 14 14 20 25 5 15 31 38 7 16 43 53 10 17 45 61 16 18 19 23 4 19 42 47 5 20 17 24 7 21 44 62 18 22 51 65 14 23 35 45 10 24 56 60 4 25 25 32 7 26 37 42 5 27 18 26 8 28 32 34 2 29 41 52 11 30 34 36 2 31 46 58 12 32 69 82 13 33 23 29 6 34 33 45 12 35 60 85 25 36 58 83 25 37 74 78 4 38 27 33 6 39 27 32 5 40 56 82 26 Mean 38.07 47.07 9.10 SD 14.00 17.95 6.70 COV (%) 36.77 38.14 73.64 *FC: First crack impact resistance **UC: Ultimate crack impact resistance
INPB 16 33 23 6 20 18 14 19 13 43 11 22 18 16 16 18 19 39 36 28 13 79 17 54 38 6 14 19 49 47 7 24 13 4 7 19 22 15 8 11 22.35 15.46 69.20
ACCEPTED MANUSCRIPT
Table 6. Some statistical results obtained from implemented tests on the specimens Max.
Range***
51.41 57.24 61.48
45.86 50.75 48.95
55.89 62.41 67.79
10.03 11.66 18.84
0.999 0.646 0.999
0.879 0.507 0.978
5.07 5.60 6.25
4.05 4.32 4.84
5.72 6.25 7.17
1.67 1.93 2.33
0.999 0.999 0.999
0.983 0.965 0.997
17 16 24
74 123 141
57 107 117
0.999 0.999 0.685
0.438 0.27 0.096
23 23 30
85 147 180
62 124 150
0.271 0.834 0.568
0.062 0.260 0.053
42.55 63.26 86.52 52.91 79.87 112.14
AC C
EP
KS*: Kolmogorov–Smirnov AD**: Anderson–Darling Range***: Difference between the minimum and maximum
SC
RI PT
Min.
TE D
The compressive strength GFRC0.25 49.00 GFRC0.75 54.42 GFRC1.25 56.84 The flexural strength GFRC0.25 4.69 GFRC0.75 5.14 GFRC1.25 5.71 The first crack impact resistance GFRC0.25 33.59 GFRC0.75 48.68 GFRC1.25 65.42 The ultimate crack impact resistance GFRC0.25 41.44 GFRC0.75 62.22 GFRC1.25 84.50
p-value AD**
Upper 0.95% CI of mean
M AN U
Lower 0.95% CI of mean
KS*
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Table 7. Number of replications required to retain the error under a specified limit at 90% and 95% level of confidence GFRC0.25 GFRC0.75 GFRC1.25 Error (%) FC UC FC UC FC UC 90% level of confidence <10 23 24 25 28 31 32 <15 10 11 12 13 14 15 <20 6 6 7 7 8 8 <25 4 4 4 5 5 6 <30 3 3 3 4 4 4 <35 2 2 3 3 3 3 <40 2 2 2 2 2 2 <45 2 2 2 2 2 2 <50 1 1 1 2 2 2 95% level of confidence <10 37 40 41 45 51 53 <15 17 18 19 20 23 24 <20 10 10 11 12 13 14 <25 6 7 7 8 9 9 <30 5 5 5 5 6 6 <35 3 4 4 4 5 5 <40 3 3 3 3 4 4 <45 2 2 3 3 3 3 <50 2 2 2 2 3 3
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
Fig 1. Distribution of used fine aggregate
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
Fig 2. Used waste GFRP sheets as reinforcement of self-compacting concrete
ACCEPTED MANUSCRIPT
SC
RI PT
Distributed recycled GFRP fibers
AC C
EP
TE D
M AN U
a) b) Fig 3. a) Slump flow test for reinforced self-compacting concrete; b) Microscopic images to indicate the homogeneity of fibers in the reinforced self-compacting concrete with 1.25% recycled GFRP fibers
EP
TE D
M AN U
a)
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
b) Fig 4. a) Some prismatic beams and the adopted test setup for execution of FPB test; b) Disks and apparatus of impact test
ACCEPTED MANUSCRIPT
T500=1.6949Vf +4.7966 2
M AN U
a)
SC
RI PT
R =0.963
TE D
Tv=2.1017Vf+10.068 2 R =0.857
b)
AC C
EP
Fig 5. Effect of different glass fiber percentage contents on: a) Slump flow and T500; b) TV
SC
RI PT
ACCEPTED MANUSCRIPT
b)
d)
AC C
EP
c)
TE D
M AN U
a)
e) Fig 6. Effect of different glass fiber percentage contents on: a) Compressive strength; b) Flexural strength; c) The first crack impact resistance; d) The ultimate crack impact resistance; e) INPB
GFRC0.75
GFRC1.25
M AN U
SC
GFRC0.25
RI PT
ACCEPTED MANUSCRIPT
TE D
Distributed recycled GFRP fibers
AC C
.
EP
Fig 7. Failed disks under drop weight impact loading
ACCEPTED MANUSCRIPT
Rupture of glass fibers
TE D
M AN U
SC
Cement hydrated products
RI PT
Cement hydrated products
AC C
EP
Fig 8. SEM images from predominated failure modes of recycled glass fiber
Rupture of glass fibers
SC
RI PT
ACCEPTED MANUSCRIPT
ffr (MPa)
b) Flexural strength
FC (Blow) c)
TE D
M AN U
a) Compressive strength
The first crack impact resistance
UC (Blow) d) The ultimate crack impact resistance
AC C
EP
Fig 9. The histogram of the impact resistance and mechanical properties of specimens GFRC0.25
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ffr (MPa)
b) Flexural strength
EP
TE D
a) Compressive strength
c)
AC C
FC (Blow)
The first crack impact resistance
UC (Blow) d) The ultimate crack impact resistance
Fig 10. Probability distribution of the impact resistance and mechanical properties of specimens GFRC0.25
ACCEPTED MANUSCRIPT
ffr=0.1561fc-2.9533 R2=0.974
M AN U
a)
SC
ffr=0.1599fc-3.5576 R2=0.949
RI PT
ffr=0.1151fc-0.8257 R2=0.985
INPB=0.8616fc-38.825 R2=0.941
TE D
INPB=0.9451fc-43.82 R2=0.881
EP
INPB=0.5587fc-23.902 R2=0.923
AC C
b) Fig 11. a) Compressive strength vs. flexural strength; b) Compressive strength vs. INPB
UC=2.5231fc-93.39 R2=0.958
M AN U
a)
SC
FC=2.3901fc-92.96 R2=0.958
RI PT
ACCEPTED MANUSCRIPT
UC=3.9359fc-169.91 R2=0.916
TE D
FC=2.9957fc-128.87 R2=0.869
AC C
EP
b)
UC=2.767fc-115.07 R2=0.974
FC=3.2559fc-130.19 R2=0.987
c) Fig 12. Compressive strength vs. the first or ultimate crack impact resistance for: a) GFRC0.25; b) GFRC0.75; c) GFRC1.25
M AN U
a)
SC
UC=1.258×FC-0.7225 R2=0.965
RI PT
ACCEPTED MANUSCRIPT
TE D
UC=1.205×FC+3.5999 R2=0.989
AC C
EP
b)
UC=1.3043×FC-0.7664 R2=0.991
c) Fig 13. The first crack impact resistance vs. the ultimate crack impact resistance
ACCEPTED MANUSCRIPT Highlights
The impact resistance and mechanical properties of reinforced SCC with rGFRP. The compressive and flexural strength of reinforced SCC follow normal distribution.
AC C
EP
TE D
M AN U
SC
RI PT
The impact resistance and mechanical properties were correlated linearly.