Study on Flexural Fatigue Properties of POM Fiber Airport Pavement Concrete

Polyoxymethylene (POM) fiber is a new polymer fiber with the potential to improve the performance of airport pavement concrete. The effect of POM fiber on the flexural fatigue properties of concrete is an important issue in its application for airport pavement concrete. In this study, four-point flexural fatigue experiments were conducted using ordinary performance concrete (OPC) and POM fiber airport pavement concrete (PFAPC) with fiber volume contents of 0.6% and 1.2%, at four stress levels, to examine the flexural fatigue characteristics of these materials. A two-parameter Weibull distribution test of flexural fatigue life was performed, after examining the change in flexural fatigue deformation using the cycle ratio (n/N). A flexural fatigue life equation was then constructed considering various failure probabilities (survival rate). The results show that POM fiber had no discernible impact on the static load strength of airport pavement concrete, and the difference between PFAPC and OPC in terms of static load strength was less than 5%. POM fiber can substantially increase the flexural fatigue deformation capacity of airport pavement concrete by almost 100%, but POM fiber had a different degree of detrimental impact on the fatigue life of airport pavement concrete compared to OPC, with a maximum decrease of 85%. The fatigue lives of OPC and PFAPC adhered to the two-parameter Weibull distribution, the single- and double-log fatigue equations considering various failure probabilities had a high fitting degree based on the two-parameter Weibull distribution, and their R2 was essentially over 0.90. The ultimate fatigue strength of PFAPC was roughly 4% lower than that of OPC. This study on the flexural fatigue properties of POM fiber airport pavement concrete has apparent research value for the extension of POM fiber to the construction of long-life airport pavements.


Introduction
Cement concrete is often used for airport pavement structures. Presently, aircraft are carrying increasingly heavy loads, and takeoffs and landings happen more frequently, so the load on airport pavement structures is becoming increasingly severe; cracking, angle loss, and plate fractures frequently occur during the course of operations [1]. As such, the performance of airport pavement concrete must be improved. During use, airport pavement concrete is subject to impact load as well as aircraft fatigue load. Concrete deterioration under fatigue stress is one of the primary causes of structural failure and a main contributor to structural durability damage [2].
Based on a trabecular three-point flexural test, Zheng et al. [3] investigated the flexural fatigue of high-strength steel-fiber polymer concrete, and discovered that the addition of 0.64 wt% steel fiber and 0.015 wt% polymer latex substantially increased concrete toughness and flexural fatigue resistance. However, few test results were available at that time. For In this study, PFAPC trabeculae with two different POM fiber contents, along with ordinary performance concrete (OPC) trabeculae, were subjected to four-point flexural fatigue testing. Under various load levels, the flexural fatigue deformation and fatigue life were measured. The fatigue life of these materials was statistically investigated, the relevant fatigue equation was created, and the flexural fatigue deformation properties were examined under various cycles. Then, the final fatigue strength was calculated.

Raw Materials and Mix Proportion
The mix ratio of OPC was calculated in accordance with the specifications for airport cement concrete pavement design (MH/T 5004-2010) [22]. To increase the workability and early performance of a dry and rigid concrete mixture for airport pavement concrete, P·O 42.5 cement, 4.75~16 and 16~26.5 mm double-graded gravel, and machine-made sand with a fineness modulus of 3.1 were used. F-type I low-calcium fly ash, S95-grade blast furnace slag powder, and polycarboxylic acid high performance water reducing agent were also added (fly ash replaced 16% of the cement mass, slag powder replaced 11% of the cement mass, and water reducing agent comprised 0.5% of the binding material's total mass). The POM fiber was obtained from Chongqing Yuntianhua Tiamjuxincai Co., Ltd. (Chongqing, China); Table 2 displays its physical and mechanical properties. Figure 1 depicts the shape of POM fiber; the microscopic images were obtained using a ZEISS microscope (Carl Zeiss AG, Oberkochen, Baden-Württemberg, Germany). POM fibers were introduced to the fresh concrete at 0.6 and 1.2% of the OPC volume. After being thoroughly agitated, the fresh concrete was packed into a 100 × 100 × 400 mm beam for vibration compaction and mold testing. The four-point flexural fatigue test was performed 24 h after the product had been poured. The product was withdrawn from the mold and placed at a temperature range of 20 ± 2 • C where the relative humidity was above 95%, for standard curing over a period of 90 days. Additionally, a 100 × 100 × 100 mm cube and a beam with the same size as the fatigue specimen were created, to test the compressive and flexural strengths at 28 and 90 days. Table 3 displays the mix proportion of the specimens.

Raw Materials and Mix Proportion
The mix ratio of OPC was calculated in accordance with the specifications for airport cement concrete pavement design (MH/T 5004-2010) [22]. To increase the workability and early performance of a dry and rigid concrete mixture for airport pavement concrete, P·O 42.5 cement, 4.75~16 and 16~26.5 mm double-graded gravel, and machine-made sand with a fineness modulus of 3.1 were used. F-type I low-calcium fly ash, S95-grade blast furnace slag powder, and polycarboxylic acid high performance water reducing agent were also added (fly ash replaced 16% of the cement mass, slag powder replaced 11% of the cement mass, and water reducing agent comprised 0.5% of the binding material's total mass). The POM fiber was obtained from Chongqing Yuntianhua Tiamjuxincai Co., Ltd. (Chongqing, China); Table 2 displays its physical and mechanical properties. Figure 1 depicts the shape of POM fiber; the microscopic images were obtained using a ZEISS microscope (Carl Zeiss AG, Oberkochen, Baden-Württemberg, Germany). POM fibers were introduced to the fresh concrete at 0.6 and 1.2% of the OPC volume. After being thoroughly agitated, the fresh concrete was packed into a 100 × 100 × 400 mm beam for vibration compaction and mold testing. The four-point flexural fatigue test was performed 24 h after the product had been poured. The product was withdrawn from the mold and placed at a temperature range of 20 ± 2 °C where the relative humidity was above 95%, for standard curing over a period of 90 days. Additionally, a 100 × 100 × 100 mm cube and a beam with the same size as the fatigue specimen were created, to test the compressive and flexural strengths at 28 and 90 days. Table 3 displays the mix proportion of the specimens.

Test Method
Before the test, the specimen surface was polished using a hand-held concrete grinder after being removed from the curing room. A high-pressure pneumatic sprayer was used to spray a thin layer of epoxy resin evenly on the surface of specimens, to reduce the test error produced by the unevenness of the surface. The static load strength was established prior to the fatigue test. The flexural and fatigue tests were performed using an MTS 810 test machine (MTS Systems Corporation, Eden Prairie, MN, USA), and the compression test using hydraulic pressure testing equipment with a loading rate of 18 kN/s. To produce roughly one-dimensional stress conditions in the center of the specimen, the purely flexural part of the four-point loading method was used. The specimen was loaded at a rate of 1 mm/min until damaged was observed, with loading points spaced 100 mm apart. To choose the upper and lower limits of loading that corresponded to the stress levels in the fatigue test, the failure strength was determined by taking the average value of three recordings.
In this study, repeated loading at a frequency of 10 Hz was used, corresponding to a commercial airliner taxiing at a speed of 20 km/h on a runway [23], with constant amplitude sine wave loading during the fatigue test loading to simulate the actual waveform of airport surface stress [24]. The cyclic characteristic value (R) of fatigue load for concrete roads, bridges, and other structures is 0.1 (the minimum fatigue loading stress is 10% of the maximum stress) [25]. The stress scale (S) was 0.65, 0.70, 0.75, and 0.80. Additionally, two strain gauges and a displacement sensor were placed at the bottom of each specimen, along the symmetry axis of the pure bending section. A dynamic strain tester (DH 15202, Donghua Testing Technology Co., Ltd., Jingjiang, Jiangsu, China) was attached to the test instrument. The recorded loading load, tensile strain in the pure bending region at the bottom of the beam prior to breaking, and midspan deflection deformation were simultaneously recorded during the loading procedure. To ensure that the epoxy glue on the contact surface was compacted and in complete contact with the loading point of the testing apparatus, preloading (10 kN) was repeatedly performed before the fatigue test began. Figure 2 displays a photograph of the loading setup. Figure 3 displays an experimental flowchart. Five specimens were chosen for each loading condition in the fatigue test, and the data were deleted and supplied with significant deviations to confirm the validity of the data on the fatigue life of the five specimens, as previously suggested [26]. neously recorded during the loading procedure. To ensure that the epoxy glue on the contact surface was compacted and in complete contact with the loading point of the testing apparatus, preloading (10 kN) was repeatedly performed before the fatigue test began. Figure 2 displays a photograph of the loading setup. Figure 3 displays an experimental flowchart. Five specimens were chosen for each loading condition in the fatigue test, and the data were deleted and supplied with significant deviations to confirm the validity of the data on the fatigue life of the five specimens, as previously suggested [26].      Table 4 shows that the compressive strength of PFAPC with 0.6 and 1.2% POM fibers in comparison with OPC ranged from −9.7 to −8.2% at 28 days, and 8.9 to 10.2% at 90 days, respectively. The flexural strength at 28 days was between −3 and 1.3%, whereas the flexural strength at 90 days was between −2 and 2.1%. The POM fiber had no discernible impact on the ability of the concrete airport pavement to withstand static loads.

Flexural Fatigue Deformation
In this study, the average bottom tensile strain and maximum midspan deflection of specimens were employed to describe the fatigue deformation at various stress levels. In  Table 4 displays the static characteristics of each group of concrete specimens at the test ages.  Table 4 shows that the compressive strength of PFAPC with 0.6 and 1.2% POM fibers in comparison with OPC ranged from −9.7 to −8.2% at 28 days, and 8.9 to 10.2% at 90 days, respectively. The flexural strength at 28 days was between −3 and 1.3%, whereas the flexural strength at 90 days was between −2 and 2.1%. The POM fiber had no discernible impact on the ability of the concrete airport pavement to withstand static loads.

Flexural Fatigue Deformation
In this study, the average bottom tensile strain and maximum midspan deflection of specimens were employed to describe the fatigue deformation at various stress levels. In Figure 4, the maximum midspan deflection and average tensile strain at the bottom are shown on the ordinate; the variation curves of specimen deformation were generated on the abscissa with the cyclic ratio n/N (the ratio of current cycle times to fatigue life) based on 11 data points. The maximum midspan deflection and average tensile strain at 1, 5, 10, 20, 50, 70, 80, 90, 95, 99.5 and 100% of specimen fatigue life at each stress level represent the 11 selected data points [27].  The maximum midspan deflection and average tensile strain of OPC and PFAPC both increased as the cycle ratio increased, and their deformation characteristics were broadly categorized into three stages [4,5,27,28]: development, stable, and failure stages, as shown in Figure 4. The interior microcracks of the material started to increase under fatigue stress during the development stage (often before a 20% cyclic ratio). The maximum midspan deflection and average tensile strain rapidly increased with increasing cycle ratio, as a result of the cracks in the weak region of the matrix. Fewer fissures were present at this point, and the material began to suffer interior damage. The maximum midspan deflection and average tensile strain gradually and consistently increased, and new fractures started to emerge in the material during the stable period (generally at a cycle ratio between 20 and 80%). The damage to the material expanded at a stable rate. In the failure stage (mainly at a cycle ratio greater than 80%), sharp increases were observed in specimen instability, internal damage across the maximum deflection, and the average tensile strain. The stress level and the average tensile strain specimen failure (when the average tensile strain was 100%) increased with increasing POM fiber content, increasing average maximum deflection, and increasing failure in specimen cross tensile strain. In this phase, the main crack matrix was internal, but due to the influence of POM fiber bridge cracks, the front of the specimen could absorb more destructive energy, delaying specimen destruction. According to Liu et al. [5], the main crack width matrix is responsible for the stage of accelerated fatigue deformation. PFAPC is tougher under fatigue load than OPC. At the 0.80 stress level, the maximum midspan deflection and average tensile strain of the 0.6% PFAPC group increased by 9.5 and 43.4%, respectively; those of the 1.2% PFAPC group increased by 37.9 and 89.2%, respectively. Under the 0.65 stress level the The maximum midspan deflection and average tensile strain of OPC and PFAPC both increased as the cycle ratio increased, and their deformation characteristics were broadly categorized into three stages [4,5,27,28]: development, stable, and failure stages, as shown in Figure 4. The interior microcracks of the material started to increase under fatigue stress during the development stage (often before a 20% cyclic ratio). The maximum midspan deflection and average tensile strain rapidly increased with increasing cycle ratio, as a result of the cracks in the weak region of the matrix. Fewer fissures were present at this point, and the material began to suffer interior damage. The maximum midspan deflection and average tensile strain gradually and consistently increased, and new fractures started to emerge in the material during the stable period (generally at a cycle ratio between 20% and 80%). The damage to the material expanded at a stable rate. In the failure stage (mainly at a cycle ratio greater than 80%), sharp increases were observed in specimen instability, internal damage across the maximum deflection, and the average tensile strain. The stress level and the average tensile strain specimen failure (when the average tensile strain was 100%) increased with increasing POM fiber content, increasing average maximum deflection, and increasing failure in specimen cross tensile strain. In this phase, the main crack matrix was internal, but due to the influence of POM fiber bridge cracks, the front of the specimen could absorb more destructive energy, delaying specimen destruction. According to Liu et al. [5], the main crack width matrix is responsible for the stage of accelerated fatigue deformation. PFAPC is tougher under fatigue load than OPC. At the 0.80 stress level, the maximum midspan deflection and average tensile strain of the 0.6% PFAPC group increased by 9.5% and 43.4%, respectively; those of the 1.2% PFAPC group increased by 37.9% and 89.2%, respectively. Under the 0.65 stress level the maximum midspan deflection and average tensile strain of the 0.6% dosage group increased by 80% and 35.3%, respectively, whereas those of the 1.2% dosage group increased by 101.2% and 36.5%, respectively. The average tensile strain growth rate of OPC and PFAPC accelerated as the stress level increased, particularly in the failure stage, as shown in Figure 4b. Additionally, as evidenced by fatigue deformation test findings for the majority of fiber-reinforced concrete materials [4,5,27], differing stress levels have no appreciable impact on the fatigue deformation of specimens within a group.

Fatigue Life Probability Distribution
Weibull distribution was used for assessing the distribution of the fatigue life of concrete materials [29]. The two-parameter Weibull distribution equation can be condensed into Equation (1) [30,31]: Then, Equation (1) can be written as: where b is the Weibull shape parameter (slope parameter), p is the reliability, N is the fatigue life, and N a is the characteristic life parameter. As Equation (2) is a linear equation, each Weibull parameter can be directly determined from the fitting line. Two-parameter Weibull distribution of experimental data is considered valid if regression analysis demonstrates a strong linear relationship between Y and X. Table 5 displays the fatigue life for each stress level, where the data for fatigue life (n) from each stress level are organized from small to large, starting with serial number i. Equation (3) can be used to compute the reliability p related to the fatigue life N: The averaged fatigue life values from Table 5 are depicted as a histogram in Figure 5. The averaged fatigue life values from Table 5 are depicted as a histogram in Figure  5.     According to the information in Table 5, the ln[ln(1/p)]-lnN curves of OPC and PFAPC were developed with varying fiber contents under various stress levels, as shown in Figure 6.  Figure 6 shows that the OPC and PFAPC test data fitting lines were linear at different stress levels. A strong, statistically significant linear relationship was found between ln[ln(1/p)] and lnN, and the correlation coefficient R 2 was greater than 0.9. The fatigue life could be described by a two-parameter Weibull distribution. Table 6 provides a summary of the Weibull distribution parameters of OPC and PFAPC at various stress levels, as shown in Figure 6.   Figure 6 shows that the OPC and PFAPC test data fitting lines were linear at different stress levels. A strong, statistically significant linear relationship was found between ln[ln(1/p)] and lnN, and the correlation coefficient R 2 was greater than 0.9. The fatigue life could be described by a two-parameter Weibull distribution. Table 6 provides a summary of the Weibull distribution parameters of OPC and PFAPC at various stress levels, as shown in Figure 6.

Test of Fitting Degree of Fatigue Life Probability Distribution
The Kolmogorov-Smirnov (K-S) method was used to test the degree of fatigue life fit, to confirm the utility of the two-parameter Weibull distribution for assessing the fatigue life distribution of OPC and PFAPC. Small sample sizes (less than 20) are better-suited for the K-S test [32]. Equation (4) expresses the K-S technique: where N i is the fatigue life associated with serial number i; the coefficient b and the life characteristic N a were substituted into the preceding formula. Table 7 summarizes the results of the fitting test of the two-parameter Weibull distribution of OPC and PFAPC. By checking the K-S critical value table, when the sample size n was five and the significance level was 0.05, the critical value D c was 0.563, the observed statistical values D i of each group were all less than D c , and the K-S test results was accepted, which further verified that the fatigue life distribution of OPC and PFAPC followed the two-parameter Weibull distribution. The significance level was less than 5%.

Flexural Fatigue Equation
The single-logarithm (S-lgN) and double-logarithm (lgS-lgN) fatigue equations for concrete are frequently used by engineers to match fatigue performance curves. The single-logarithm fatigue equation is as follows [33]: The fatigue equation has two boundary conditions [33]: The second boundary condition of the single-log fatigue equation cannot be satisfied, preventing the determination of the fatigue properties of concrete at low stress levels (S < 0.50). The double-log fatigue equation [30] is established to prolong the fatigue curve in the direction of S→0: The test findings and the aforementioned boundary requirements can both be satisfied by this ideal form of the fatigue equation. S-N curves are traditionally created using the stress level S as the ordinate and the average fatigue life as the abscissa [4]. As illustrated in Figure 7, the standard S-N curves of OPC and PFAPC were constructed in this study using the average fatigue life in shown Figure 5.
The second boundary condition of the single-log fatigue equation cannot be satisfied, preventing the determination of the fatigue properties of concrete at low stress levels (S < 0.50). The double-log fatigue equation [30] is established to prolong the fatigue curve in the direction of S→0: The test findings and the aforementioned boundary requirements can both be satisfied by this ideal form of the fatigue equation. S-N curves are traditionally created using the stress level S as the ordinate and the average fatigue life as the abscissa [4]. As illustrated in Figure 7, the standard S-N curves of OPC and PFAPC were constructed in this study using the average fatigue life in shown Figure 5.   To meet the safety performance requirements under cyclic loading in practical engineering, the P-S-N fatigue equations of OPC and PFAPC were established under a given failure probability in the form of single-and double-logarithm fatigue equations. The aim was to establish a direct quantitative relationship between the regression fatigue equation and failure probability F (or survival rate P = 1 − F). The fatigue lives of OPC and PFAPC both followed a two-parameter Weibull distribution, according to the analysis in Sections 3.3 and 3.4. Equation (10) can be used to compute the fatigue life N f under various failure probabilities F: The coefficients b and N a obtained in Table 6 were substituted into Equation (10), and the fatigue lives N f (equivalent fatigue life) of OPC and PFAPC were calculated with given failure probability F, as shown in Table 8. The data in Table 8 were regressed using Equations (7) and (9). Table 9 provides a summary of the regression coefficients A, B, lga, and b of the single-and double-log fatigue equations, which correspond to various failure probabilities F. Table 9 demonstrates that, regardless of the shape of the single-or double-logarithm form, the correlation R 2 between OPC and PFAPC rapidly declined as the failure probability increased. In the single-logarithmic form of fitting, the correlation of other groups under different failure probabilities was over 0.90, and the correlation of the PFAPC group was above 0.97, with the exception of a slightly lower correlation in OPC when the failure probability was 0.40-0.50. With the exception of a marginally lower correlation in OPC when the failure probability was 0.20-0.50, the correlations of all other groups in the doublelog fitting were over 0.90, whereas those of the PFAPC group were above 0.96. This showed that when the failure probability F is considered, the equivalent fatigue life N f of the PFAPC obeyed the two-parameter Weibull distribution with high accuracy. Additionally, the corresponding linear relationship between the single-log fatigue equation and the doublelog fatigue equation was essentially established, with the degree of fit for the single-log fatigue equation being slightly higher than that of the double-log equation. This is in line with the results of the fitting correlations between the average fatigue life corresponding to the single-and double-log fatigue equations. Failure probability F had little impact on regression coefficients B and b, and can be ignored, similar to the majority of the fatigue test results for fiber-reinforced concrete. The single-and double-logarithm fatigue equations of airport pavement concrete, considering failure probability F, can be determined by using the average value of B and b as the general result, as shown below: These equations serve as a guide for forecasting the fatigue life of PFAPC with various failure probabilities under various stress levels, when the above formulae are combined with the regression coefficients under various failure probabilities as shown in Table 9. Flexural fatigue testing of concrete generally yields two S-N curves with high reference values. The first is the P-S-N curve, which corresponds to a survival rate of 50%, or a failure probability of 50%; the maximum fatigue strength of a material can be determined through this curve. The second is the corresponding survival-rate P-S-N curve; P is 95%, which means that F is 0.05 in terms of failure probability. The ultimate strength of the material under conditional fatigue can be determined by this curve, and this value can serve as a general guide for structural design [34,35].  According to a previous study [3], a concrete specimen has an unlimited life if it is not harmed after 2 × 10 6 cycles of loading. The ultimate fatigue strength of material generally refers to the maximum fatigue stress that the material can withstand under a certain number of cycles, and is usually expressed in the form of static flexural strength percentage in practical application [4,7,24]. Researchers studying the ultimate fatigue strength of concrete materials typically use a cyclic foundation of 2 × 10 6 fatigue cycles. The OPC and PFAPC ultimate fatigue strength (upper limit of stress level) and conditional ultimate fatigue strength corresponding to a 2 × 10 6 fatigue life under two survival rates was calculated using Figures 8 and 9, as well as the established single-and double-log fatigue equations, as shown in Table 10.   According to a previous study [3], a concrete specimen has an unlimited life if it is not harmed after 2 × 10 6 cycles of loading. The ultimate fatigue strength of material generally refers to the maximum fatigue stress that the material can withstand under a certain number of cycles, and is usually expressed in the form of static flexural strength percentage in practical application [4,7,24]. Researchers studying the ultimate fatigue strength of concrete materials typically use a cyclic foundation of 2 × 10 6 fatigue cycles. The OPC and PFAPC ultimate fatigue strength (upper limit of stress level) and conditional ultimate fatigue strength corresponding to a 2 × 10 6 fatigue life under two survival rates was calculated using Figures 8 and 9, as well as the established single-and double-log fatigue equations, as shown in Table 10.  According to a previous study [3], a concrete specimen has an unlimited life if it is not harmed after 2 × 10 6 cycles of loading. The ultimate fatigue strength of material generally refers to the maximum fatigue stress that the material can withstand under a certain number of cycles, and is usually expressed in the form of static flexural strength percentage in practical application [4,7,24]. Researchers studying the ultimate fatigue strength of concrete materials typically use a cyclic foundation of 2 × 10 6 fatigue cycles. The OPC and PFAPC ultimate fatigue strength (upper limit of stress level) and conditional ultimate fatigue strength corresponding to a 2 × 10 6 fatigue life under two survival rates was calculated using Figures 8 and 9, as well as the established single-and double-log fatigue equations, as shown in Table 10. No difference in the ultimate fatigue strength was estimated by using the singlelogarithm and double-logarithm form equations, as shown in the calculation results in Table 10. The ultimate fatigue strengths of OPC, PFAPC-0.6, and PFAPC-1.2 with a survival rate of 50% were found to be 0.68 f r (where f r is the static flexural strength of OPC and PFAPC), 0.66 f r , and 0.65 f r , respectively, equating to a 2 × 10 6 fatigue life. The conditional ultimate fatigue strengths of OPC, PFAPC-0.6, and PFAPC-1.2 corresponding to a 2 × 10 6 fatigue life were calculated as 0.65 f r , 0.64 f r , and 0.63 f r , respectively, when considering a survival rate of 95%. The average fatigue life of the material in the fatigue test was indicated to be larger than 2 × 10 6 times when the loading stress amplitude was less than or equal to the ultimate fatigue strength, suggesting no fatigue failure. When the survival rate was 50%, the fatigue strength of concrete with POM fibers at 0.6 and 1.2% volume decreased by 2.9 and 4.4%, respectively; when the survival rate was 95%, the fatigue strength fell by 1.5 and 3.1%, respectively. The addition of POM fiber reduced the flexural fatigue performance of airport pavement concrete to a certain extent.

Future Recommended Research
The conclusion shows that POM fibers have different degrees of negative effects on the fatigue life of airport pavement concrete. This is different from the results from other studies of polymer-fiber-reinforced concrete. These findings were obtained under the premise of ensuring the validity of the test results. It is hoped that this conclusion will be taken seriously by more colleagues and more research will be conducted to clarify why these negative effects occur, and how they should be addressed.