Effect of Short Fibers on Fracture Properties of Epoxy-Based Polymer Concrete Exposed to High Temperatures

Recently, polymer concrete (PC) has been widely used in many civil engineering applications. PC shows superiority in major physical, mechanical, and fracture properties comparing to ordinary Portland cement concrete. Despite many suitable characteristics of thermosetting resins related to processing, the thermal resistance of polymer concrete composite is relatively low. This study aims to investigate the effect of incorporating short fibers on mechanical and fracture properties of PC under different ranges of high temperatures. Short carbon and polypropylene fibers were added randomly at a rate of 1 and 2% by the total weight of the PC composite. The exposure temperatures cycles were ranged between 23 to 250 °C. Various tests were conducted including flexure strength, elastic modulus, toughness, tensile crack opening, density, and porosity to evaluate the effect of addition of short fibers on fracture properties of PC. The results show that the inclusion of short fiber lead to an increase in the load carrying capacity of PC by an average of 24% and limits the crack propagation. On the other hand, the enhancement of fracture properties of based PC containing short fibers is vanished at high temperature (250 °C), but still more efficient than ordinary cement concrete. This work could lead to broader applications of polymer concrete exposed to high temperatures.


Introduction
Polymer concrete (PC), one of the three types of polymer concrete type, is formed by completely substituting Portland cement with polymer binders. The two other types are polymer-impregnated concrete (PIC), which consists of a hardened Portland cement concrete impregnated with a low-viscosity monomer polymerized in situ, and the second one is the polymer cement concrete (PCC) which is produced by incorporating a monomer or polymer in a Portland cement concrete mix and polymerized after placing concrete [1].
The polymer binder used in PC is acting through a polymerization process to bond the solid particles of concrete [2,3]. PC, whether reinforced or non-reinforced, shows superiority in major physical, mechanical, and fracture properties when compared to ordinary Portland cement. These properties include high mechanical strength, rapid hardening, and short curing time, which are very beneficial for precast element production or partitions with high toughness. [4]. In addition, PC is attributed with fast demolding, minimum cracking, higher mechanical strength, good bonding to old substrates, better abrasion resistance, chemical resistance, and less maintenance in comparison to Portland cement concrete [5,6]. This explains the quick development of using PC in many civil engineering applications [7].
Various types of resins have been used in the production of polymer concrete such as epoxy [8,9], unsaturated polyester [10,11], acrylics and vinyl-ester, poly acrylate, polypropylene, furan-based resin, and other polymers [12,13]. However, the formed composites incorporating these materials are brittle and the majority of their properties are very sensitive to long exposure at high temperatures [14][15][16][17]. This is because the polymers acting as Two kinds of fibers (carbon and polypropylene) were chosen to be embedded into the PC's mixes (Table 1). A carbon fiber marketed as Aceca ® ECO-H6 was supplied by Torayca Company, Japan. These fibers (having cylindrical shapes) are chopped into short lengths in the interest to help dispersion in viscous fluids. The second type of fiber, labeled under the name of DUOMIX FIRE ® (M6), was provided by Bekaert (Belgium) as fine monofilaments polypropylene. These fibers have a tubular shape and come in the form of clusters which can be easily dispersed during mixing. The main disadvantage of these types of fibers was lowering the workability of the mix for the same volume of polymer.
Sand graded from 0 to 4 mm and a gravel graded from 4 to 10 mm were used as fine and coarse aggregates, respectively. The specific gravity of the sand was 2.47, whereas the specific gravity of gravel was 2.53 g/cm 3 . Preliminary investigations based on the compressible packing model (CPM) was conducted to obtain the proportions of aggregates that guaranteed the maximum compactness of solid grains while minimizing the amount of polymer needed to wet fully the aggregates and fill gaps between. Portland cement CEM I 52.5 N having a specific density of 3.13 g/cm 3 and a specific surface equal to 3590 cm 2 /g as evaluated via Blaine apparatus was also used in this study.

Sample Preparation
In this study, four PC mixes with fibers reinforcements and 13% of polymer, by weight of PC, were prepared. The PC samples were prepared according to a process described in the work of Elalaoui et al. [13].
The content of polymer was kept constant for the all mixes. This content was selected as the optimal amount in the basis of previous studies [13,25,[32][33][34][35]. The fibers were added at the rate of 1% and 2%, by the total weight of the polymer. The PC compositions are listed in Table 2. The formed PC mixes are named as PC-Xp, where Xp% denotes the type of reinforcement (C for carbon and P for polypropylene) and p referring to the mass fraction of the introduced fibers in term of percentage. The PC was casted according to a well-defined process as described in a previous published study [13]. Ordinary Portland cement concrete (OCC) samples were also casted by mixing the same aggregates proportions for comparison purposes. The OCC mix constituent is shown in Table 2.

Testing Method
Several tests were conducted including compression and flexure strength, elastic modulus determined using the ultrasonic pulse velocity method (UPV), toughness, tensile crack opening, density, and porosity (using mercury intrusion porosimetry, MIP) to evaluate the effect of addition of short fibers on fracture properties of PC. The principle of MIP is to inject a no-wetting liquid such as mercury inside the pores of a small sample (10 cm 2 as maximum) under vacuum conditions and under pressure ranging from 14 to 414 MPa.
A uniaxial compression tests were carried out using cylinders sample (50 mm × 100 mm). The effect of adding fibers on the flexure strength, toughness, and mechanical residual properties of PC were investigated by performing three-point bending test beam (50 mm × 50 mm × 305 mm). These tests were conducted in a displacement-control mode with a constant rate of 1.0 mm/min and 1.25 mm/min for bending and compression, respectively, as per the procedure stated in RILEM PCM8 1995 [36].
The midpoint deflection of beams was evaluated by the mean of a linear variable differential transformer (LVTD) rested on a steel L-shape bracket attached to specimens at mid-height as shown in the figure below ( Figure 1). The content of polymer was kept constant for the all mixes. This content was selected as the optimal amount in the basis of previous studies [13,25,[32][33][34][35]. The fibers were added at the rate of 1% and 2%, by the total weight of the polymer. The PC compositions are listed in Table 2. The formed PC mixes are named as PC-Xp, where Xp% denotes the type of reinforcement (C for carbon and P for polypropylene) and p referring to the mass fraction of the introduced fibers in term of percentage. The PC was casted according to a welldefined process as described in a previous published study [13]. Ordinary Portland cement concrete (OCC) samples were also casted by mixing the same aggregates proportions for comparison purposes. The OCC mix constituent is shown in Table 2.

Testing Method
Several tests were conducted including compression and flexure strength, elastic modulus determined using the ultrasonic pulse velocity method (UPV), toughness, tensile crack opening, density, and porosity (using mercury intrusion porosimetry, MIP) to evaluate the effect of addition of short fibers on fracture properties of PC. The principle of MIP is to inject a no-wetting liquid such as mercury inside the pores of a small sample (10 cm 2 as maximum) under vacuum conditions and under pressure ranging from 14 to 414 MPa.
A uniaxial compression tests were carried out using cylinders sample (50 mm × 100 mm). The effect of adding fibers on the flexure strength, toughness, and mechanical residual properties of PC were investigated by performing three-point bending test beam (50 mm × 50 mm × 305 mm). These tests were conducted in a displacement-control mode with a constant rate of 1.0 mm/min and 1.25 mm/min for bending and compression, respectively, as per the procedure stated in RILEM PCM8 1995 [36].
The midpoint deflection of beams was evaluated by the mean of a linear variable differential transformer (LVTD) rested on a steel L-shape bracket attached to specimens at mid-height as shown in the figure below ( Figure 1).   (Figure 2). The mid-span deflection was recorded using an LVDT placed at the left side at the bottom of the specimens. The dimensions of sample and central U-shape notch properties are itemized in Table 3. Concrete toughness was evaluated throughout pre-notched beams using a 250 kN closed loop INSTRON machine. The crack mouth opening displacement (CMOD) was measured by virtue of clip-gauge attached to a knife-edges installed at the bottom of the beams ( Figure 2). The mid-span deflection was recorded using an LVDT placed at the left side at the bottom of the specimens. The dimensions of sample and central U-shape notch properties are itemized in Table 3. For the OCC mix, beam samples of 70 mm × 70 mm × 280 mm for bending test and cylinder samples of 150 mm × 300 mm for compression sample were casted. The OCC beams samples were loaded at rate of 0.05 MPa/s, whereas the cylinder samples was loaded at 0.5 MPa/s. The fracture toughness values of PC in mode-I, KIC, and the fracture energy GF were calculated using a two fracture parameters model [37,38] which appears to give rather realistic prediction of concrete fracture behavior. An actual crack is replaced by an equivalent fictitious crack [39].
The effective crack length and the peak load, are required to determine KIc using linear elastic fracture mechanics, LEFM. This value of KIc was found to be independent of the specimen size [40].
The stress intensity factor is given by means of Equation (1): The critical crack length is presented by Equation (3) a c = a C f 2 ( ) and α a (3)  For the OCC mix, beam samples of 70 mm × 70 mm × 280 mm for bending test and cylinder samples of 150 mm × 300 mm for compression sample were casted. The OCC beams samples were loaded at rate of 0.05 MPa/s, whereas the cylinder samples was loaded at 0.5 MPa/s. The fracture toughness values of PC in mode-I, KIC, and the fracture energy GF were calculated using a two fracture parameters model [37,38] which appears to give rather realistic prediction of concrete fracture behavior. An actual crack is replaced by an equivalent fictitious crack [39].
The effective crack length and the peak load, are required to determine K Ic using linear elastic fracture mechanics, LEFM. This value of K Ic was found to be independent of the specimen size [40].
The stress intensity factor is given by means of Equation (1): where σ NC = 3F.S 2b.d 2 and α = a c +d 0 d+d 0 The critical crack length is presented by Equation (3) The geometrical parameters S, b, d, a 0 and d 0 are described in Figure 3. The geometrical parameters S, b, d, a0 and d0 are described in Figure 3. The fracture energy is calculated according to the following relation as postulated by the following LEFM equation: The prepared specimens were heated at a rate of 0.5 °C/min until target temperature was reached according to the predefined protocol. The temperature was maintained for 3 h before being decreases with the same heating rate till samples reached the ambient temperature being kept for 24 h until the specimens were tested. To ensure that the setting temperature (from 100 °C to 250 °C) is compatible with specimen temperature, two k-type thermocouples mounted on data logger were attached to one side and in the middle of samples for a continuous recording (Figure 4). A set of four specimens was allocated for each test type. The tests were conducted on concretes allowed to cure for 7 days at prescribed room temperature for both unreinforced and reinforced PC and 28 days age for OCC.
The setting and recorded temperatures were close and only a difference not exceeding ±3 °C was reported ( Figure 5). The fracture energy is calculated according to the following relation as postulated by the following LEFM equation: The prepared specimens were heated at a rate of 0.5 • C/min until target temperature was reached according to the predefined protocol. The temperature was maintained for 3 h before being decreases with the same heating rate till samples reached the ambient temperature being kept for 24 h until the specimens were tested. To ensure that the setting temperature (from 100 • C to 250 • C) is compatible with specimen temperature, two k-type thermocouples mounted on data logger were attached to one side and in the middle of samples for a continuous recording (Figure 4). The geometrical parameters S, b, d, a0 and d0 are described in Figure 3. The fracture energy is calculated according to the following relation as postulated by the following LEFM equation: The prepared specimens were heated at a rate of 0.5 °C/min until target temperature was reached according to the predefined protocol. The temperature was maintained for 3 h before being decreases with the same heating rate till samples reached the ambient temperature being kept for 24 h until the specimens were tested. To ensure that the setting temperature (from 100 °C to 250 °C) is compatible with specimen temperature, two k-type thermocouples mounted on data logger were attached to one side and in the middle of samples for a continuous recording (Figure 4). A set of four specimens was allocated for each test type. The tests were conducted on concretes allowed to cure for 7 days at prescribed room temperature for both unreinforced and reinforced PC and 28 days age for OCC.
The setting and recorded temperatures were close and only a difference not exceeding ±3 °C was reported ( Figure 5). A set of four specimens was allocated for each test type. The tests were conducted on concretes allowed to cure for 7 days at prescribed room temperature for both unreinforced and reinforced PC and 28 days age for OCC.
The setting and recorded temperatures were close and only a difference not exceeding ±3 • C was reported ( Figure 5). Polymers 2023, 15, x 7 of 16

Fibers Adding Effects on the Density and Porosity Properties of PC
The effect of addition of short fibers on the density PC mixed is shown in Table 4. The results reveals that the inclusion of short fibers have a slight effect on the density of PC concrete. The density of the studied mixes ranged from 2.16 g/cm 3 to 2.23 g/cm 3 . In fact, a marginal decrease was observed for the entire mixes. The same trend can be observed for the carbon group of mixes as the density decreased with higher fiber dosages. The maximum difference in densities recorded over the mixes was 3.2%, which shows that incorporation of fibers with small quantities did not significantly affect the concrete density. In addition, it was noted that the density of polypropylene fibers is lower than that of polymer. This could be due to the deficiency of uniform distribution during the mixing process and thereby in concrete bulk. The pore structures (i.e., mean pore size, pore size distribution, and various pores size proportions) were characterized by means of MPI and results are presented in Table  4. The introduction of fibers leads also to a decrease in the mean most distributed pore diameter dc. The increase in the porosity of PC reinforced with polypropylene fibers can be attributed to volume expansion of the fibers during mixing which conducts to an additional porosity. In general, the addition of short fibers to polymer concrete generates difficulties of the solid particles compaction causing an increase in the total porosity and a decrease in the most distributed pore diameter, dc as shown in Figure 6. Similar phenomena have been observed by Bentur et al. [41] and Barbuta et al. [42]. Although, the porosity of PC mixes increases with fiber content, but is still significantly lower than that of OCC samples.

Fibers Adding Effects on the Density and Porosity Properties of PC
The effect of addition of short fibers on the density PC mixed is shown in Table 4. The results reveals that the inclusion of short fibers have a slight effect on the density of PC concrete. The density of the studied mixes ranged from 2.16 g/cm 3 to 2.23 g/cm 3 . In fact, a marginal decrease was observed for the entire mixes. The same trend can be observed for the carbon group of mixes as the density decreased with higher fiber dosages. The maximum difference in densities recorded over the mixes was 3.2%, which shows that incorporation of fibers with small quantities did not significantly affect the concrete density. In addition, it was noted that the density of polypropylene fibers is lower than that of polymer. This could be due to the deficiency of uniform distribution during the mixing process and thereby in concrete bulk. The pore structures (i.e., mean pore size, pore size distribution, and various pores size proportions) were characterized by means of MPI and results are presented in Table 4. The introduction of fibers leads also to a decrease in the mean most distributed pore diameter dc. The increase in the porosity of PC reinforced with polypropylene fibers can be attributed to volume expansion of the fibers during mixing which conducts to an additional porosity. In general, the addition of short fibers to polymer concrete generates difficulties of the solid particles compaction causing an increase in the total porosity and a decrease in the most distributed pore diameter, d c as shown in Figure 6. Similar phenomena have been observed by Bentur et al. [41] and Barbuta et al. [42]. Although, the porosity of PC mixes increases with fiber content, but is still significantly lower than that of OCC samples.

Effects of Adding Fibers on the Modulus of Elasticity of PC Systems
The dynamic Young's modulus E d and the shear modulus are determined using UPV. The two dynamic modulus are given by the following Equations (5) and (6): The value of Poisson's ratio is given by Equation (7): In the Equation (5) to Equtaion (6), Ed represents the dynamic elasticity modulus (MPa), VL is the compressive P-wave velocity, VT is shear wave velocity (km/s), ν is Poisson's ratio, and  is the density (kg/m 3 ).
The results of the experimental program revealed that the static elastic modulus (Es) decreased slightly as the fiber content increased (Figure 7a,b). This marginal variation is the result of two balanced phenomena: an increase due to the high elastic modulus of the carbon fibers and decrease due to the increase in porosity and the anisotropic feature. By increasing the fraction of fibers, the growth of porosity outweighs the contribution of fibers. The lowest rigidity is observed for PC-P1 because polypropylene fibers have a low elastic modulus similar to that of the epoxy binder. Moreover, PC-P1 has higher values of the total porosity compared to control sample. In the same context, Reis and Kumar [10,43] in their research paper confirmed that adding carbon and glass fibers do not improve the compressive elastic modulus of the composites but in opposite slight decrease was observed while carbon fiber reinforcements are incorporated.

Effects of Adding Fibers on the Modulus of Elasticity of PC Systems
The dynamic Young's modulus E d and the shear modulus µ d are determined using UPV. The two dynamic modulus are given by the following Equations (5) and (6): The value of Poisson's ratio is given by Equation (7): In the Equation (5) to Equtaion (6), E d represents the dynamic elasticity modulus (MPa), V L is the compressive P-wave velocity, V T is shear wave velocity (km/s), ν is Poisson's ratio, and ρ is the density (kg/m 3 ).
The results of the experimental program revealed that the static elastic modulus (Es) decreased slightly as the fiber content increased (Figure 7a,b). This marginal variation is the result of two balanced phenomena: an increase due to the high elastic modulus of the carbon fibers and decrease due to the increase in porosity and the anisotropic feature. By increasing the fraction of fibers, the growth of porosity outweighs the contribution of fibers. The lowest rigidity is observed for PC-P1 because polypropylene fibers have a low elastic modulus similar to that of the epoxy binder. Moreover, PC-P1 has higher values of the total porosity compared to control sample. In the same context, Reis and Kumar [10,43] in their research paper confirmed that adding carbon and glass fibers do not improve the compressive elastic modulus of the composites but in opposite slight decrease was observed while carbon fiber reinforcements are incorporated. By adding fibers to PC, all mixes showed indistinctly an increase in the modulus with sample age and being stabilized after a period of 5 to 7 days (Figure 8). This can be attributed to the progress of the curing reaction taken end at the 7 days as maximum leading to the formation of polymeric epoxy structures [16]. It should be mentioned that the modulus of elasticity records for unreinforced PC displayed are in accordance with the values described in literature for epoxy-based PC [11,44]. By adding fibers to PC, all mixes showed indistinctly an increase in the modulus with sample age and being stabilized after a period of 5 to 7 days (Figure 8). This can be attributed to the progress of the curing reaction taken end at the 7 days as maximum leading to the formation of polymeric epoxy structures [16]. It should be mentioned that the modulus of elasticity records for unreinforced PC displayed are in accordance with the values described in literature for epoxy-based PC [11,44].

Fibers Adding Effects on the Mechanical Properties of PC
The results of flexural and compressive strength of PC mixes are presented in Figure  9a,b. The composite systems exhibit lower flexural and compressive characteristics when reinforced with polypropylene and carbon fibers, regardless of the type of reinforcements.

Fibers Adding Effects on the Mechanical Properties of PC
The results of flexural and compressive strength of PC mixes are presented in Figure 9a,b. The composite systems exhibit lower flexural and compressive characteristics when reinforced with polypropylene and carbon fibers, regardless of the type of reinforcements. In fact, it was reported that polypropylene fibers as example show weak binder-fiber contacts and this decrease the performance of the final product, regardless of the binder type [45].

Fibers Adding Effects on the Mechanical Properties of PC
The results of flexural and compressive strength of PC mixes are presented in Figure  9a,b. The composite systems exhibit lower flexural and compressive characteristics when reinforced with polypropylene and carbon fibers, regardless of the type of reinforcements. In fact, it was reported that polypropylene fibers as example show weak binder-fiber contacts and this decrease the performance of the final product, regardless of the binder type [45].
However, the results show that flexural strength of PC made with carbon fibers increased with the increase in fibers content (Figure 9a). In average, flexural strengths increased 20% approximately for 2% fiber content, and compressive strength decreased 25.9% for the same fiber content, comparing to PC-C1.  However, the results show that flexural strength of PC made with carbon fibers increased with the increase in fibers content (Figure 9a). In average, flexural strengths increased 20% approximately for 2% fiber content, and compressive strength decreased 25.9% for the same fiber content, comparing to PC-C1.
The decrease in the mechanical resistances is probably due to poor adhesion between the fibers and the matrix and to the high values of porosity. This decrease is a sign of decrease in transverse bonds in polymer, which decrease the stiffness and increase ductility of polymer concrete [46]. In fact, adding 1% of carbon fibers does not affect the compressive strength and the elastic characteristics but decreases the flexure strength. In other words, fibers do not increase PC strength but their addition to the mixture diminished the signs of brittleness behavior of unreinforced polymer concrete [20], which is translated by a change in bending behavior from quasi-linear brittle to nonlinear ductile, a nonlinearity heightened by increasing the fiber content.
It can be concluded that the addition of fibers did not accomplish the expected reinforce or at least has the same strength characteristics as unreinforced PC [17]. Hence, only the toughness of polymer concretes reinforced with carbon fibers will be considered thereafter. Figure 10a,b represent, respectively, the variation of stress intensity factor K IC and the fracture energy G F as a function of the exposure temperature for both the PC system and the OCC. It can be seen that PC is more resistant to crack propagation than OCC. By increasing the exposure temperature, the toughness is improved up to 225 • C for both concrete before declining slightly at higher temperatures (250 • C). Such a variation has been similarly observed for the flexural strengths mainly in the case of plain PC as a result of action of two coexistent and competitive phenomena: (i) post crosslinking initiated by heat and (ii) thermo-oxidative degradation of the epoxy polymer [47]. This variation is emphasized by a loss of bond between aggregates and the binder as demonstrated in a previous work [14]. concrete before declining slightly at higher temperatures (250 °C). Such a variation has been similarly observed for the flexural strengths mainly in the case of plain PC as a result of action of two coexistent and competitive phenomena: (i) post crosslinking initiated by heat and (ii) thermo-oxidative degradation of the epoxy polymer [47]. This variation is emphasized by a loss of bond between aggregates and the binder as demonstrated in a previous work [14]. The effect of carbon fiber introduction on the fracture properties of PC is studied and the results are depicted on Figure 10a,b. It can be seen that adding 1% of carbon fibers to polymer concretes results in an enhancement on its fracture properties at room temperature, but this effect is vanished by increasing the exposure temperature as a result of the decrease in bond strength in the fiber-matrix interface while rising temperature [14] added to the other types of degradation mentioned earlier. Figure 11 shows the load-CMOD graph for the tested notched beams as a function of heating temperature. The overall load-CMOD curves have presented a smooth softening curves after the peak loads which indicates that tests were conducted under a stable test regime. All curves observed show a similar general trend for all specimens.

Fibers Adding Effects on Toughness at High Temperatures
It can be seen also that for reinforced PC mixes the load carrying capacity increases until 150 °C and then decreases for higher temperatures. Moreover, a significant change in the post-peak behavior was observed, the ductile behavior becomes less brittle [48]; similar to the type of behavior observed in non-reinforced concrete beams. This ductility enhancement yields slower crack propagation, a phenomenon attributed to fracture energy increase due to the existence of a micro-cracks network. It should be highlighted that more energy dissipation is necessary for the coalescence of micro-cracks into a single macro-crack [49]. The use of fine and short fibers in composite helps to reduce their critical crack length and increase fiber-matrix interface area to intercept cracks. This means that if crack passes through fibers (being less stiff than fiber), it will encounter the resin which will control and slow crack. If this does not stop crack propagation, the next fiber can The effect of carbon fiber introduction on the fracture properties of PC is studied and the results are depicted on Figure 10a,b. It can be seen that adding 1% of carbon fibers to polymer concretes results in an enhancement on its fracture properties at room temperature, but this effect is vanished by increasing the exposure temperature as a result of the decrease in bond strength in the fiber-matrix interface while rising temperature [14] added to the other types of degradation mentioned earlier. Figure 11 shows the load-CMOD graph for the tested notched beams as a function of heating temperature. The overall load-CMOD curves have presented a smooth softening curves after the peak loads which indicates that tests were conducted under a stable test regime. All curves observed show a similar general trend for all specimens. Polymers 2023, 15, x 13 of 16 Figure 11. Load-CMOD for PC systems for different heating temperatures.

Conclusions
The effect of incorporating short carbon and polypropylene fibers combined with the exposure to high heating temperatures reaching 250 °C on the mechanical and fracture properties of epoxy polymer-based concrete is investigated in depth in this study and the following remarks were derived: • Epoxy-based concrete possesses higher mechanical properties compared to ordinary cement concrete.

•
For temperatures less than 250 °C, the epoxy polymer concrete is still more efficient than ordinary cement concrete. It can be seen also that for reinforced PC mixes the load carrying capacity increases until 150 • C and then decreases for higher temperatures. Moreover, a significant change in the post-peak behavior was observed, the ductile behavior becomes less brittle [48]; similar to the type of behavior observed in non-reinforced concrete beams. This ductility enhancement yields slower crack propagation, a phenomenon attributed to fracture energy increase due to the existence of a micro-cracks network. It should be highlighted that more energy dissipation is necessary for the coalescence of micro-cracks into a single macrocrack [49]. The use of fine and short fibers in composite helps to reduce their critical crack length and increase fiber-matrix interface area to intercept cracks. This means that if crack passes through fibers (being less stiff than fiber), it will encounter the resin which will control and slow crack. If this does not stop crack propagation, the next fiber can intercept crack and stop it. Fibered PC force cracks to follow a very devious path that needs larger amounts of energy to create new fracture surfaces.

Conclusions
The effect of incorporating short carbon and polypropylene fibers combined with the exposure to high heating temperatures reaching 250 • C on the mechanical and fracture properties of epoxy polymer-based concrete is investigated in depth in this study and the following remarks were derived: • Epoxy-based concrete possesses higher mechanical properties compared to ordinary cement concrete. • For temperatures less than 250 • C, the epoxy polymer concrete is still more efficient than ordinary cement concrete.

•
The addition of short carbon fibers content by a rate of 1% by weight to polymer concrete indicated that there was no indicative difference in the concrete density, the elastic characteristics, the compressive strength, and led to a decrease in the flexural strength. It also did not enhance the fracture properties at room temperature. • When exposed to high temperatures of up to 150 • C, the 1% fiber introduction resulted in a load-carrying capacity increase. This enhancement of fracture properties is vanished for higher heating temperatures. • Compared to the post cracking behavior at room temperature, the ductility of 1%-fibered polymer concrete increases when it is exposed to high temperatures, resulting in slower crack propagation.

•
Carbon fibers introduction at content of 2% by weight of polymer-based concrete did not improve its mechanical and fracture properties. The same was observed for polypropylene fibers used with a fraction of 1% by weight.

Recommendations
Due to complexity and co-existence of many phenomena, deeper investigations are hence needed to explain their interaction that affect the properties of reinforced fiber polymer concrete. Curing conditions of samples and their effects in the fracture and thermo-mechanical properties seems to be interesting to be investigated and evaluated to build a complete and consistent knowledge about polymer concrete exposed to extreme in-service conditions.