Effect of Polypropylene Fiber on the Flexural Strength Properties of Lightweight Foam Mixed Soil

Civil Engineering Department, Faculty of Engineering, King Mongkut’s University of Technology !onburi, !ung Khru, Bangkok, !ailand Soil Engineering Research Center, Department of Civil Engineering, King Mongkut’s University of Technology North Bangkok, Bangkok, !ailand Construction and Building Materials Research Center, Department of Civil Engineering, King Mongkut’s University of Technology North Bangkok, Bangkok, !ailand Sustainable Infrastructure Research and Development Center, Department of Civil Engineering, Khon Kaen University, Khon Kaen, !ailand


Introduction
An improvement in the properties of soil is necessary due to constraints imposed by constructions in problematic areas such as areas with low-strength and high-compressibility soil. Two main categories of soil improvement include improvement of the in situ foundation soil and improvement of the filled soil. Lightweight cement-treated soil is well recognized among geotechnical engineers and has been used in many civil engineering applications, such as embankment material on soft ground, backfill material, filling material, and filter material [1][2][3]. is improvement technique was first used in Norway for road embankment structures in 1972 [4], and later, the volume of such constructions increased rapidly. Among the various types of lightweight cement-treated soil, lightweight foam mixed soil (LFS) is one of the most popular materials in road embankment construction. e main strategy is to add cement content and foam in an amount that increases the strength and decreases the density of the mixture, respectively. e mechanical properties of LFS under various loading conditions and mixing components have been studied [2,[5][6][7][8][9]. e influence of cement content, curing time, density, and proportion of foaming agent on the strength behavior of lightweight foam mixed soils has been studied by laboratory tests. With good performance, low price, and well-controlled production, the utilization of LFS has been extended to various applications, particularly the maintenance of road structure and approach to bridges under operation. For main roads where operations cannot be suspended for a long time, precast LFS is preferred. e LFS can be produced in a plant nearby or even at the construction site; and the precast LFS from nearby plants is transported to the construction location. However, with high cement content and low densities, the structure of LFS consists of a cement-soil matrix and a large amount of both pores and voids (larger). As a result, the LFS exhibits extremely brittle behavior [6,7]. is brittle behavior is becoming an important consideration, particularly in the use of precast LFS. During transportation and installation, cracking and chipping can easily occur.
is imparts unfavorable properties to the repaired road.
Extensive literature has reported that the addition of various types of randomly discrete fibers into a cementitious matrix is well recognized and widely practiced for various structural applications. Fiber inclusion leads to improvements in the mechanical behavior of cement-treated soils under shear [10][11][12][13][14][15][16][17][18][19], compression [20][21][22][23], and flexural [24][25][26] test conditions. Yetimoglu and Salbas [27] stated that the shear strength and stress-strain behaviors of sands reinforced with a random distribution of discontinuous fibers depend on the fiber content, aspect ratio, and surface friction between the fiber and cemented matrix. Temel et al. [28] concluded that the fiber content has a strong effect on the bearing capacity, stiffness, and ductility of sand fills reinforced with fiber inclusion. Estabragh et al. [29] demonstrated that the mechanical behavior of soil-cement-fiber mixtures is a function of the fiber content and curing time. According to the results of an experimental study by Casagrande et al. [30], an increase in the fiber length and content results in an increase in the peak strength of fiberreinforced bentonite materials. Fiber inclusion has been proven to be efficient in cement-soil structures. However, for materials with cement-soil structures with large voids, such as the LFS, little investigation has been conducted to reveal the effects of fiber inclusion on their mechanical behavior. e objective of this study was to investigate the flexural characteristics of LFS that are improved by polypropylene fiber. To achieve this aim, the entire testing program for the laboratory tests includes simple physical property tests and flexural performance tests. e polypropylene fiber is prepared at three different volume fractions: 0.5%, 0.75%, and 1%. e desired unit weights of LFS are 0.8, 1.0, and 1.2 (g/ cm 3 ) prepared by varying the quantity of foam. ree different cement contents (100, 150, and 200 kg/m 3 ) of soil are used. e curing period of specimens is fixed at 28 days. Finally, the overall rating of flexural performance is applied to provide useful insight for evaluating the efficiency of the fiber content on the flexural behavior of the LFS material.

Materials
In this study, soft clay classified as CH in the Unified Soil Classification System (USCS) from the campus of King Mongkut's University of Technology onburi, ailand, was used in all tests. e fiber material used in the present study was polypropylene fiber, which is widely employed in fiber-reinforced cemented soil [24,25,[31][32][33][34][35][36][37][38]. e polypropylene fiber was prepared with a length of 58 mm at three different volume fractions: 0.5%, 0.75%, and 1%. Type I Portland cement was used in this study. e foam was made by mixing foaming material and air in an air foam generator. e required density was 0.05 t/m 3 . e foaming material was made by diluting the foaming agent with water, and the proportion of the mixture (foaming agent: water) is 1 : 19, as suggested by PWRI [39]. e basic physical properties of the materials used in this study are shown in Table 1.

Method of Remolding the Clay.
e clay used herein was remolded until it became mud. e water contents were set as the liquid limit (LL) of approximately 1.5-3.0 LL. e purpose of the remolding water content is to increase the workability and flowability. To prevent the bleeding of aircement-treated soil, a flow value of 450 mm from the flow test is required [39]. Prior to the addition of cement content and foam, natural soil was mixed with the associated addition of water. e remolding clay water content (w * ) is hereinafter defined as the water content of the remolded clay prior to the addition of cement content and foam. e amount of water added to a wet clay sample to obtain the desired remolding water content was calculated by the following fundamental equation: where ΔW W , W T , w * , and w 0 are the additional weight of water to be added, total weight of prepared original untreated clay sample, required remolding clay water content, and natural water content of the clay sample, respectively. e samples of base clay with the required amount of additional water were placed inside a portable mechanical soil mixer and allowed to mix thoroughly for a few hours. When the mixed clay was uniformly remolded, the water content was closer to the desired remolding water content within ±1%, and the sample was then ready for admixture mixing. e clay sample or mud was placed in a humid room overnight to maintain the water content of the sample.

Mixing Procedures.
e preparation of the specimens is shown in Figure 1.
e specimen was 100 mm wide, 100 mm high, and 350 mm long. Prior to admixing with foam, the prepared remolded clay sample at a particular remolding water content was mixed with cement and polypropylene fiber inside the soil mixer for approximately 6 minutes until the mixture became a slurry. A constant remolding water content of 300% was used in all tests. Subsequently, air foam was added and mixed with the slurry. e desired density of air-cement-treated soil was adjusted to achieve the target density within a range of ±0.03 g/cm 3 , and the range of owability values for the ow test was between 16 and 20 cm [39]. Since the foam has water, the overall water content of the clay-water-cement-foam just after the time of mixing was the total remolding water plus the water in the foam. For the soil cement, the total clay water content (C w ) and the total clay water content (C w ) of air-cement-treated soil can be adapted by the relation in equation (2). e overall water content in the mixture is hereinafter called the mixing water content or the total clay water content (C w ). e total clay water content (C w ) is de ned as follows: where C w , w * , w/w F , and F w are the total clay water content (or mixing water content) of the clay-water-cement-y ash-

Advances in Materials Science and Engineering
foam paste (%) from the dry weight of soil only, the remolding clay water content (%) before mixing the cement and foam, the water-foam material ratio, and the desired foam (%), which is defined as the percentage ratio of the weight of foam to the dry weight of soil. e overall aircement-treated soil preparation for the entire study consisted of 300% remolding water content at the initial state in all mixtures and at controlled densities of 0.8, 1.0, and 1.2 g/ cm 3 . e cement content, A w , is defined as 100, 150, and 200 (kg) in 1 m 3 of air foam-cement-admixed clay. e details of the experimental program are presented in Table 2. e specimen with a density of 0.8 g/cm 3 , cement content of 100 kg/m 3 , and fiber content of 1% is represented by 0.8C100F1 as the test code in this study. According to Miller and Rifai [40], polypropylene has a relatively high melting point, low thermal and electrical conductivity, and a high ignition point, with a specific gravity of 0.91. It is also a hydrophobic and chemically inert material that does not absorb or react with soil moisture or leachate. In this study, the polypropylene fiber was prepared with a length of 58 mm and used in the study at three different volume fractions: 0.5%, 0.75%, and 1%.

Flexural Performance Test.
After curing for 28 days, the dimensions of the specimens were measured, and all variations in width, height, and length were satisfactory. e flexural performance test was conducted in accordance with the ASTM C1609 standard [41]. is standard is a specification governing the testing of fiber-reinforced concrete beams conducted on a beam with third-point loading ( Figure 2). e test is run at a specified net deflection rate to a net deflection of L/600 (where L is the support span distance). It can then be run at a higher specified net deflection rate until the specified endpoint.
To better understand the relative influence of the fiber content, in this study, the term equivalent flexural strength ratio was used to describe the efficiency of fiber inclusion on the engineering characteristics of LFS, for both deflection softening and deflection hardening. is index provides the determination of specimen toughness based on the area under the load-deflection curve up to the prescribed deflection, as depicted in Figure 3. e flexural strength (f) can be expressed as follows: where P is the load (kN) corresponding to L, b, and d which are span length, width, and depth of the specimen, respectively. For the deflection softening (gray line), the first peak load (P 1 ) is defined as the peak load (P p ) and corresponds to the deflection value δ p . For numerical purposes, P 1 is appropriately defined as the first crack point [25,42,43]. To assess the ductility of LFS, the ductility index (DI) in the approach proposed by Nematollahi et al. [43] is used and referred to as DI � (δ p /δ 1 ), where δ p and δ 1 are the deflections of the specimen at the peak and first peak, respectively. is index is productive when either the first peak strength or the overall peak strength is clearly observed.
To evaluate the flexural toughness of LFS with different fiber contents, the flexural toughness (T D 150 ) is calculated by the total area under the load-deflection curve corresponding to specified deflection of L/150 as shown in Figure 3. After obtaining the toughness capacity of LFS, the equivalent flexural strength (R D T,150 ) ratio indicates the flexural performance or ability to persist a large deformation of LFS upon axial compression until 10%. is concept is applied herein to quantify the flexural strength, and R D T,150 can be determined as follows: Zone I: R D T,150 < 100%; P 1 � P p ; low toughness Zone II: R D T,150 > 100%; P 1 < P p ; high toughness

Results and Discussion
As shown in Figure 4(a), the increase in cement content contributes to the increase in the exural strength of LFS. e exural strength also increased as the density increased. e unreinforced specimen typically showed brittle failure. With those unreinforced specimens, after reaching peak stress, failure occurred suddenly and was accompanied by a large single crack development under the sample, as depicted in Figure 4(b). e initial crack was observed from a very early stage in the load-de ection response of the entire specimen. Figure 4(b) presents the plot of δ p and cement content for unreinforced specimens, where δ p was found to decrease with a decrease in density and an increase in the cement content. It is possible that the LFS material tends to be more brittle due to the addition of void and cement content. All LFS specimens exhibit a low energy absorption ability. Fiber addition to LFS is known to change the behavior from brittle to a more ductile one. With the contribution of ber, LFS can withstand strains compared to unreinforced specimens.
e ber inclusion is more e cient in highdensity specimens, with few exceptions. It can be inferred that all test specimens have similar exural strength behavior with respect to the inclusion of ber content. e rate of exural strength decreases with increasing ber content. e maximum value for the rst peak was measured to be 630 kPa within 1.0C200F1. One feature observed from Figure 5 is that the f p values are higher than f 1 for the high ber inclusion specimens, for instance, 0.8C100F1, 0.8C200F1, and 1.0C100F1. For the unreinforced specimens, a sudden failure is observed. According to all test results, the exural characteristics of LFS were enhanced by ber addition, both of the peak and residual exural strength (f s ). As observed, although the reinforced samples show a signi cant increase in the exural performance, the ber inclusion does not improve their sti ness. In the case of cement content 100 kg/m 3 , the ber content of 0.5% samples show that their sti ness is lower than that of the unreinforced samples. In Figures 5(b), 5(c), 5(e), 5(f ), 5(h), and 5(i), the sti ness of reinforced samples does not appear to depend on the amount of ber inclusion.     Figure 6(c). According to Kheli et al. [44], f 1 directly corresponds to matrix failure. erefore, the presence of ber may a ect the postpeak behaviors, as shown in the improvement compared to the unreinforced specimens. e ratio may indicate the e ect of ber existence, which can be used to improve the ductility or deformability of LFS. It was found that, by the inclusion of ber, there was a remarkable e ect on the lightly cemented samples. e peak strength ratio is almost the same when the cement content is increased. Figure 6(d) shows that a more pronounced residual strength is obtained in the case of high ber content. For the unreinforced specimens, the residual exural strength was found to be 0 kPa (sudden failure), as has been previously mentioned.
In terms of the ductility index (DI (δ p /δ 1 )), it is evident that the DI values of lowest cement content were higher than those with cement content 150 and 200 kg/m 3 , as depicted in Figure 7(a). e increase in ductility index for the mixture with cement content 100 kg/m 3 is remarkably increased with the density. As expected, almost all samples with high cement content exhibited DI 1. It is likely that ber does not contribute to increasing DI when the cement content is 150 and 200 kg/m 3 , which produced softening behavior. To re ect the toughness behavior, Figure 7(b) shows the variations of T D 150 corresponding to a given set of testing programs. e results illustrate that more ber in all mixtures results in a higher tendency of T D 150 . However, the e ect of cement content has an insigni cant impact on the variation of T D 150 . Figure 7(c) shows the relationship between R D T,150 and ber content at di erent densities and cement contents. ese gures illustrate that a more pronounced toughness behavior is obtained with decreasing cement content and increasing density specimens. A high   Advances in Materials Science and Engineering 7 toughness behavior (R D T,150 > 100%) is found (specimens 1.0C100F1.0, 1.2C100F0.5, and 1.2C100F1.0), as evident from Figures 5(g) and 5(h). Referring to Figure 5, it can be seen that by reinforcing, di erent peak and residual strengths (de ection hardening and de ection softening) were obtained. e trend of de ection hardening is noticeably larger for the high-density and low-cement content specimens (e.g., 1.2C100F1.0, 1.2C100F0.75, and 1.2C150F0.5) due to the dominant e ects of ber inclusion. It can be concluded that the e ciency of ber has a positive in uence on the toughness behavior with increasing density and decreasing cement content, barring a few exceptions.
To evaluate the test results, the concept of overall performance was applied to incorporate the performance of LFS reinforcement on exural behaviors. e concept of overall performance was developed by Jamsawang et al. [26] as an appropriate parameter to evaluate the exural performance of compacted cement-ber-sand (CCFS). To achieve the best overall performance rating, six subperformance values were used: (1) rst peak strength (f 1 ), (2) peak strength (f p ), (3) residual strength (f s ), (4) ductility index (DI), (5) exural toughness at de ection of L/150 (T D 150 ), and (6) equivalent exural strength ratio (R D T,150 ). After obtaining each subperformance value, the average subperformance value at each density and cement content, as shown in Table 3, is calculated using the following equation: average subperformance value sum of performance value number of fiber contents × 100.
Based on the obtained values of the average subperformance value, the comparative score average subperformance value can be calculated with a maximum score of 100 for each cement content using the following formula: comparative score average sub performance value best average subperformance value for each cement content × 100.
As depicted in Table 4, for comparison, the values of f 1 for subperformance with a cement content of 100 kg/m 3 for    percentage of overall performance � sum of score of subperformance sum of full score of subperformance × 100.
e percentage of overall performance is used to account for a specific application in the design process. From the present test results, with the addition of polypropylene fiber, the overall performance values increase with increasing density for cement contents of 150 and 200 kg/m 3 , respectively. However, for the lowest cement content, the fiber inclusion seems to produce a strong effectiveness at a density of 1.0 g/cm 3 . e overall performance rating was divided into seven levels: (1) excellent, (2) very good, (3) good, (4) fairly good, (5) fair, (6) poor, and (7) very poor. Because there was no standard to define these levels, the common concept of significance in geotechnical engineering was adopted, as listed in Table 5. Table 6 shows the ranks of all mixing proportions based on the overall performance of LFS (%). It can be observed that only sample 1.0C100 provides excellent overall performance, whereas samples 1.2C150 and 1.2C100 show very good performance. In addition, samples 1.2C200, 1.0C200, and 1.0C150 achieved good performance, whereas fair performances were achieved with samples 0.8C150 and 0.8C100. Moreover, some applications may require a specific subperformance rather than an overall performance. As the density increases, there are many more existing contacts of fiber surfaces, products of cement hydration, and particles of soil and interfaces. is results in high forces developed through the interfacial mechanical interactions, and thus, the effectiveness of the fibers is greater [26,45]. Table 6 can help a designer select the optimal mixing proportion for a specific application.
According to all test results, the deformation characteristics of LFS were enhanced by fiber addition, especially in terms of the loss of postpeak stress. To explain this phenomenon, when local cracks begin to form in the specimen containing fibers, the fibers act as bridges between crack openings, resisting crack development, and the stresses are transferred across by means of fibers in a behavior known as "fiber bridging" [46][47][48]. e fibers extend from one surface   is results in high forces developed through the interfacial mechanical interactions, and thus, the e ectiveness of the bers is greater. e loss of postpeak is directly related to the propagation of local cracks, which start forming at the peak stress state. e presence of bers is then likely to obstruct crack propagation, particularly due to the bers that are perpendicularly aligned to the crack propagation direction. ese bers bear the induced shear stress, and this induced shear stress around the interface begins to exceed the interfacial bond strength at crack formation. e bond force decreases due to the sliding of bers along the interfacial zone, resulting in a decrease in the rate of softening. As shown in Figure 8, visual observations indicate that the ber impedes the further development of cracking, resulting in double cracks and smaller crack widths for specimens 1.0C100F0.75 and 1.2C200F1.0, respectively.

Conclusions
After an experimental investigation of the exural behaviors of ber-reinforced lightweight foam mixed soil, polypropylene  fiber was found to have a significant influence on the flexural behaviors of LFS, and the following conclusions can be drawn: (i) e flexural strength testing results showed that the contribution of polypropylene fiber to flexural strengths increases the first peak, peak, and residual flexural strengths. e polypropylene fiber improved the flexural strength substantially when compared with the unreinforced specimen. For the fiber inclusion specimens, the first peak, peak, and residual flexural strengths were noted to increase with an increase in density and cement content. (ii) By increasing the fiber content, the fiber inclusion does not improve the stiffness of LFS. e stiffness of reinforced samples is almost constant and does not appear to be dependent on the amount of fiber inclusion. (iii) In terms of ductility, the fiber inclusion increased the ductility index significantly for the lowest cement content specimens. For the mixture with a cement content of 100 kg/m 3 , the ductility index increases with the density. (iv) For the toughness behavior, with the contribution of fiber, LFS can withstand larger strains compared to unreinforced specimens. e fiber inclusion is more efficient with an increase in density and decrease in cement content specimens, which resulted in a high toughness behavior (R D T,150 > 100%). (v) Based on the conception of overall performance, with the addition of polypropylene fiber, the overall performance values increase with increasing density for cement contents of 150 and 200 kg/m 3 . However, for the lowest cement content, the fiber inclusion seems to produce a strong effectiveness at a density of 1.0 g/cm 3 . (vi) e failure mode of LFS with the inclusion of fiber impedes the further development of cracking, which results in noticeable multiple cracks and/or smaller crack widths.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest regarding the publication of this paper.