Effect of the Addition of Dispersed Reinforcement on the Resilient Modulus of Slightly Cemented Non-Cohesive Soil

Subbase and subgrade are the foundation of the pavement structure. It has to provide a short-term surface for traffic and a suitable platform for the placement and compaction of the high-quality asphalt layer above. In addition, it has to work as a load-bearing system for the completed pavement during the required design period (Brown & Selig, 1991).

A measure of material stiffness, which is a parameter used in the mechanistic-empirical pavement design procedure, is the resilient modulus (M_r). It characterises the non-linearity of subgrade or subbase granular soil response to the traffic load. Repeated loading causes recoverable and irrecoverable components of deformation. The resilient behaviour of unbound materials used in pavement layers is dependent on a few factors, described in the literature (Lekarp et al., 2000; Kim & Kim, 2007): grain size distribution, maximum grain size, the content of fines, particle shape and aggregate type, density, moisture content, compaction method, stress level, number of load cycles, load duration and load sequence and freeze–thaw cycles.

The M_r is the parameter expressed as the ratio of applied cyclic deviator stress σ_d to recoverable (elastic/resilient) axial strain during unloading ɛ_r (Seed et al., 1962), as shown in Eq. (1): (1) Mr=σd/εr {M_r} = {\sigma _d}/{\varepsilon _r} and it is defined as an elastic modulus of the material under repeated loads. The M_r can be determined in the triaxial cell with the ability to the application of cyclic loading, named in the literature as repeated load triaxial laboratory test (RLT). During the test, certain confining pressure, axial stress and a number of cycles are applied in sequences according to different regulations.

To avoid large deformations caused by traffic load, pavement construction materials are improved by adding various additives. A lot of research was conducted on materials that might be used as additives, such as cement (Saxena et al., 2010; Ismail et al. 2014), lime (Ahmed et al., 2020; Yin et al., 2022), fly ash from different sources (Yilmaz, 2015; Mahvash et al., 2018; Noaman et al., 2022; Wang et al., 2022; Zimar et. al, 2022), and various types of fibres (Consoli et al., 2003; Yetimoğlu et al., 2003; Khattak & Alrashidi, 2006). The use of fibres as a reinforcement increases stiffness (Festugato et al. 2023), load-bearing capacity (Yetimoğlu et al., 2005; Rashid et al., 2017), the tensile strength of soils (Consoli et al., 2013; Li et al., 2014) and reduces potential liquefaction (Ibraim et al., 2010).

As the literature concluded (Consoli et al., 1998; Park, 2011; Hamidi & Hooresfand, 2013), cement addition increased the stiffness of the soil and its brittleness, while the fibres added to sandy soil changed its failure behaviour from brittle to more ductile one.

In tests conducted on compacted sand reinforced with 50 mm long polypropylene fibres, the soil monotonic mechanical parameters were improved (Festugato et al., 2023). Fibres also had an impact on the cyclic behaviour of dense sand evidenced by an increase in cycles to failure. Fibres incorporated in cemented sandy soil caused an increase in unconfined compression strength (Consoli et al., 2010, 2013; Khodabandehlou et al., 2023).

The aim of this article is to investigate the influence of the dispersed reinforcement – polypropylene fibres added in different lengths and amounts – on the resilient response of the coarse material with the slight cement addition (1.5%), used as a base or subbase in road pavement construction. Polypropylene fibres incorporated in the tested soil amended with cement had lengths of 12, 18 and 40 mm and were added in amounts of 0%, 0.2% and 0.3% by weight of dry soil.

The authors concluded from their previous research that it is not possible to provide the resilient modulus tests on non-cohesive soil without fines because the cyclic triaxial test destroys unbound non-cohesive soil (Zabielska-Adamska et al., 2021, 2023). The soil was tested as hydraulically bound with a 1.5% cement addition, chosen as the minimum amount that could improve the resilient characteristics of the tested soil. The fibre addition higher than 0.3% was not considered because it decreased the static mechanical properties of the tested soil (Zabielska-Adamska et al., 2023).

Laboratory tests were conducted on gravelly sand grSa (EN ISO 14688-1), which particle size distribution according to the EN 933-1 standard is illustrated in Figure 1. The tested non-cohesive soil is poorly graded (EN ISO 14688-2). Its coefficient of uniformity, C_U, and coefficient of curvature, C_C, were 5.04 and 0.66, respectively. As authors characterised material in earlier studies (Zabielska-Adamska, et al. 2021), the tested gravelly sand is a glaciofluvial soil (Pleistocene), collected in the Sokółka district in north-eastern Poland.

Figure 1:

The gravelly sand (grSa) was tested alone, as soil stabilised with the addition of 1.5% Portland cement 42.5R, and with an addition of polypropylene fibres with variable lengths 12 mm, 18 mm and 40 mm, and diameters 0.030 mm, 0.034 mm and 0.90 mm, respectively. Fibres were added in the amount of 0.2 and 0.3% to the dry mass of the tested soil. In the case of fibres 40 mm in length, only an addition of 0.2% was used. The images of the fibres are presented in Figure 2.

Figure 2:

It can be observed (Fig. 2) that the fibres with 40 mm long have a different structure than 12 and 18 mm reinforcement. It was also visible after mixing with the soil, that the mix with 12 and 18 mm was more homogenous than the soil with the reinforcement 40 mm long, which can be observed in Figure 3. Mixtures of soil and polypropylene fibres were mixed using a laboratory mechanical stirrer for a time of 4.5 minutes, which is of great importance for the homogeneity of the tested samples.

Figure 3:

The compaction curves of gravelly sand and sand stabilised with cement are presented in Figure 4. The exemplary compaction curves of soil with fibres 18 mm long are presented in Figure 5. The specific density of the tested samples and their compaction parameters, optimum moisture content (w_opt) and maximum dry density (ρ_{d max}), are summarised in Table 1. The compaction parameters were determined by the standard Proctor (SP) and modified Proctor (MP) tests according to the EN 13286-2 standard.

Figure 4:

Figure 5:

Based on Figures 4 and 5 and Table 1, it can be concluded that the cement addition has an impact on the maximum dry density of the tested material. In both compaction methods, standard and modified, the ρ_{d max} increased after cement addition, while the w_opt remained practically the same. In the case of soil with fibres addition, the optimum moisture content was slightly reduced after the addition of 0.2% and 0.3% of 18 mm fibres. The maximum dry density value increased significantly after fibre addition, hence the ρ_{d max} values for soil with 0.2% and 0.3% of fibres do not differ much. The specific dry density slightly increases with the addition of cement to the mixture, whereas the addition of polypropylene fibres commonly does not affect the specific dry density because of low fibre mass.

The main test of resilient modulus M_r was performed as repeated load triaxial test RLT using the triaxial cell designed to conduct the cyclic test in accordance with AASHTO T307 standard. In this type of test, a sample is subjected to 16 sequences of cyclic loading, where the first one – sequence 0 – is the conditioning of the sample (with a number of 500 to 1000 cycles), and 15 other sequences (with 100 cycles), with the different combinations of confining pressures and applied cyclic loads, depending on the cycle. The confining pressure ranged from 20.7 to 137.9 kPa and the maximum axial stress ranged from 20.7 to 275.8 kPa. The machine used repeated cycles of the haversine-shaped load pulse, where the load pulse lasted 0.1 s and the rest period 0.9 s. In the used cyclic triaxial apparatus, the confining pressure and axial load were put on pneumatically. Axial deformations of the tested sample were measured using two linear variable differential transformers (LVDTs). The data obtained during the test was collected by the computer. The results, M_r values, are calculated to 5 last cycles from every 15 sequences. The sum of the last 5 modulus readings is recorded as the resilient modulus of that sequence.

Figure 6 is an example of the cycles to which the samples were subjected during the resilient modulus test. The maximum applied vertical stress on the soil sample is σ₁, which is the sum of σ₃ (confining pressure) and σ_d (deviator stress).

Figure 6:

Samples for resilient modulus tests were prepared in a bipartite mould of 70 mm in diameter and a height-to-diameter ratio of 2. At first, dry components: soil, cement and polypropylene fibres were mixed using the laboratory mechanical stirrer, and next the required amount of water was added to get optimum moisture content. The samples were compacted dynamically in three layers to the values of maximum dry density from the SP and MP compaction tests. The M_r tests were conducted on specimens after 7 and 28 days of curing in constant temperature and humidity.

In Figure 7 one of the initial cycles of loading/unloading is shown, where the principle of the resilient modulus test is presented. In one cycle, the soil sample is subjected to loading and unloading and the permanent and recoverable deformations are visible. After each load application, granular material undergoes some irreversible deformation (Brown, 1996). Figure 8 presents the complete 100 cycles of the 15th sequence of loading and unloading according to AASHTO T307 standard. The reduction of permanent strain with successive load/unload cycles in a given sequence is clearly visible.

Figure 7:

Figure 8:

The data obtained from the resilient modulus M_r tests of improved gravelly sand after different curing times are presented in Table 2. The results presented as graphs for 0%, 0.2% and 0.3% of fibre contents are shown in Figures 9 and 10.

Figure 9:

Figure 10:

It can be observed (Tab. 2) that the addition of fibres can decrease the values of resilient modulus, especially in a shorter time of curing. In the previous article of the authors, the plasticisation of samples after fibre addition was proved (Zabielska-Adamska et al., 2023). Samples compacted by the SP method with fibres 18 mm and 12 mm (only 0.3%) gained lower M_r values after a longer period of curing. However generally, samples cured for 28 days achieved the highest values of the M_r, particularly in the case of samples compacted by the MP method. In the case of the MP compaction method, resilient modulus gained higher values than for the samples with the same amount of addition but compacted with the SP method. The increase of fibre content from 0.2 to 0.3% causes the increase of the M_r values, at similar densities of samples. The greatest values of the M_r were obtained for fibres of 18 mm in length, independent of the content of fibres, which is more visible at standard compaction. The lower values of the M_r were observed for fibres of 40 mm in length.

Depending on their position in the pavement structure, the structure materials are subjected to different loads from the traffic. The M_r values at the different magnitudes of traffic loading represented by confining pressures: 20.7, 34.5, 68.9, 103.4 and 137.9 kPa for samples with 18 mm long fibres were plotted in Figures 11 and 12, and for samples with 12 mm long fibres were plotted in Figures 13 and 14.

Figure 11:

Figure 12:

Figure 13:

Figure 14:

It can be concluded (Figs. 11–14) that the resilient modulus is affected by deviatoric and confining stress. The M_r value increases with rising confining pressure, but at the same confining pressure with rising maximum axial load. The same tendency was observed by Georgees et al. (2018). As stress increases, the sample becomes denser and therefore a lower recoverable strain is obtained. It is connected to the hardening behaviour of granular materials. Achievement of the highest values of the M_r for the modified compacted samples, which was observed for hydraulically stabilised samples (Zabielska-Adamska et al., 2021) is not so clear in the case of 1.5% cement-stabilised specimens with fibre addition, tested at the same confining pressure. The addition of fibres plasticises cement-stabilised samples (Zabielska-Adamska et al., 2023), which changes sample behaviour. Similar behaviour was observed in fibre-reinforced cemented sandy soil, where fibre addition changed the brittle failure behaviour to be more ductile (Sadek, 2010; Gao & Zhao, 2012). The greater values of the M_r were obtained for 0.3% of fibres compared to 0.2% of fibres for the same level of stress.

Based on the test results of gravelly sand compacted with cement and dispersed reinforcement, the following conclusion can be drawn:

Addition of fibres to gravelly sand stabilised with 1.5% cement decreases the resilient modulus, independent of the percentage amount of fibres and their length.