Introduction

Seasonally frozen zones cover 53.5% of China's territory, predominantly located at latitudes above 24°1,2,3. The Northeast region exemplifies such an area, where subgrade soil routinely undergoes high-frequency freeze–thaw (F–T) cycles, both short-term and continuous. In Northwest Jilin Province, silty sand, a common subgrade material, is characterized by low natural water content, a small plasticity index, and low shear strength4. These properties predispose the subgrade to issues such as deformation, ruts, and congestion during F–T cycles5,6,7. Consequently, enhancing the mechanical properties and F–T resistance of silty sand is crucial. Improved subgrade fillers not only mitigate road frost damage and prolong subgrade service life but also reduce road maintenance costs. Recent advancements have shown that inorganic binders such as cement, lime, and fly ash are effective in subgrade stabilization, with cement demonstrating superior enhancement8,9,10. Despite its benefits, cement-modified soil remains brittle, possessing high unconfined compression strength (UCS) but low tensile strength11. To address this, fiber-reinforced cemented soil has been developed through composite modification, combining fibers with the cemented matrix. This new material has garnered considerable interest from researchers both domestically and internationally due to its enhanced tensile strength12.

Basalt fiber (BF) is an emerging high-performance inorganic fiber that offers significant environmental benefits. It exhibits superior tensile strength, chemical stability, resistance to high temperatures, and resistance to acid and alkali corrosion compared to traditional fibers such as polypropylene and lignin13,14. Additionally, BF is more cost-effective than glass or carbon fibers due to its lower raw material consumption and the absence of additives15,16. The chemical composition of BF closely aligns with that of cement, enhancing the bond strength between them17. This compatibility has led to its widespread use in reinforcing the mechanical properties of both concrete and asphalt concrete18,19,20,21,22. Furthermore, the integration of BF into cemented soil has been shown to improve mechanical properties by strengthening particle bonds. Notably, the combined use of cement and BF outperforms the application of each material alone23. In light of these advantages, recent research has increasingly focused on BF-reinforced cemented soil.

The unconfined compressive strength test is an essential method for evaluating the effectiveness of cement stabilization and assessing the influence of various factors on the strength of cemented soil. Recognized by the American Society for Testing and Materials as a standard procedure, this test serves as a quantitative measure to assess soil improvement. Extensive research has focused on the UCS of BF-reinforced soil. Wang et al.23 found that adding basalt fibers significantly improved the UCS of cemented kaolinite clay, with the best results at a fiber content of 0.2%. This improvement exceeded that achieved with the use of single-doped fibers. Additionally, the inclusion of fibers considerably enhanced the ductility and strength properties of soft soils. Lv et al.24 noted that BF reinforcement in bio-cemented sand provided better results than either polypropylene or carbon fibers, reaching the highest UCS values. Ghanbari et al.25 reported that the introduction of either BF or cement significantly increased the UCS of peaty soils, with cement stabilization offering more substantial improvements than mere fiber reinforcement. However, samples reinforced with basalt fibers showed increased ductility.

Acoustic emission (AE) technology, a dynamic and nondestructive evaluation method, effectively captures the internal damage of materials from deformation to failure26. Widely applied in geotechnical engineering, seismic monitoring, and rock failure prediction, AE has demonstrated substantial utility27,28,29,30. Mao and Towhata31 employed AE detection to observe the single-particle fragmentation process under static conditions, noting an initial low AE signal activity that was followed by a period of inactivity and culminated in a significant increase in activity, marked by a high-frequency AE signal aggregation just before fracture failure. This progression yielded crucial insights into the initiation and propagation of internal cracks during particle crushing. Novikov et al.32 investigated the AE characteristics of frozen soil under quasi-static mechanical and cyclic thermal loads, confirming the suitability of the AE parameters and primary data processing methods used. Li et al.33 detected soil damage through accumulative ring-down counts during the unconfined compressive strength test, correlating changes in AE parameters with stress levels to characterize the damage. Additionally, Liang et al.34 observed that in sandstone, the cumulative ring count shifted from abrupt changes to gradual increases over repeated F–T cycles, indicating that AE characteristic parameters and ring counts effectively depicted the damage in F–T sandstone.

In summary, research into the mechanical properties of BF-reinforced soil has advanced to a mature stage. However, the analysis of the suboptimal mechanical properties of subgrade silty sand, especially under the specific environmental conditions of freeze–thaw (F–T) cycles, is still limited. Furthermore, numerous experimental studies have shown that acoustic emission (AE) characteristic parameters effectively outline the damage evolution process in samples. Nonetheless, the application of AE techniques in subgrade engineering within seasonally frozen zones remains sporadic. Research on the F–T damage characteristics of subgrade materials in these zones under load is scarce. Consequently, this paper aims to investigate the effects of various factors, including fiber content, fiber length, curing time, and F–T cycles, on the UCS of silty sand, cemented silty sand, fiber-reinforced silty sand, and fiber-reinforced cemented silty sand. Additionally, this study utilizes AE testing to elucidate the evolution of the F–T damage process in fiber-reinforced cemented silty sand under uniaxial compression and to provide early warnings of subgrade failure. This research intends to provide reliable reference data for the construction and disease prevention of BF-reinforced cemented silty sand subgrade in seasonally frozen zones.

Materials and test programs

Materials

This paper investigates BF-reinforced cemented silty sand. Based on previous research by our team, the optimal cement content for stabilized silty sand was established at 2%35, denoting the cement mass relative to the dry soil mass as defined by Chen et al.36. The silty sand for the subgrade was obtained from Songyuan County, Jilin Province. Adhering to the Chinese Specification JTG 3430-2020, "Test Methods of Soils for Highway Engineering" (T0115, T0116, T0118)37, this study provides a detailed analysis of the particle size distribution and primary physical properties of both silty sand and cemented silty sand, illustrated in Fig. 1 and Table 1, respectively. Portland cement type P.O 42.5, used in this study, was sourced from Changchun Yatai Cement Co., Ltd. in Jilin Province. The BF was obtained from Haining Anjie Composite Material Co., Ltd., in Zhejiang Province. The fibers were cut to lengths of 6 mm, 12 mm, and 18 mm, as depicted in Fig. 2. The properties of these fibers are detailed in Table 2.

Fig. 1
figure 1

Grain size distribution curve of silty sand.

Full size image
Table 1 The basic physical properties of silty sand and cemented silty sand.
Full size table
Fig. 2
figure 2

Basalt fibers: (a) dispersed 6 mm, (b) dispersed 12 mm, (c) dispersed 18 mm, (d) clustered 6 mm, (e) clustered 12 mm and (f) clustered 18 mm.

Full size image
Table 2 The characteristics of basalt fibers.
Full size table

As depicted in Table 1, the liquid limit of silty sand is 22.4%, the plasticity index is 8%, the optimum water content is 8.5%, and the maximum dry density is 2.03 g/cm3. According to the parameters listed, this soil is classified as low liquid limit silty sand under the Unified Soil Classification System, as per ASTM D2487-1738. For the cemented silty sand, the optimum water content increases to 9.7% and the maximum dry density to 2.04 g/cm3. These adjustments suggest that the incorporation of cement improves the water absorption capacity of silty sand.

As Table 2 showed, the diameter of fiber is 7–15 μm, the density is 2.63 g/cm3, the tensile strength, elastic modulus, and melting point of monofilament fiber are 3000–4800 MPa, 91–110 GPa, and 1050 °C, respectively, indicates that the fiber has great mechanical properties and great resistance to temperature.

Test programs

Fiber lengths in the samples were set at 0 mm, 6 mm, 12 mm, and 18 mm, increasing in increments of 6 mm. Fiber content is defined as the percentage of fiber relative to the total mass of soil and cement23. Initially, water was added to silty sand for mechanical mixing, followed by storage in an airtight container for 8 h to ensure uniform water distribution. Subsequently, fibers in the designated ratio were incorporated into the prepared soil; additional water was added to reach the optimum water content of 8.5%. The mixture was then mechanically mixed for 10 min to prepare fiber-reinforced silty sand (Group 1)39,40. Similarly, fibers and cement in the designated ratio were added to the prepared soil, with water added until the optimum water content of 9.7% was achieved, forming fiber-reinforced cemented silty sand (Group 2). This mixture was placed into a mold in three layers, each compacted 27 times using a manual method. The surface of each layer was roughened before the next was added to enhance sample integrity. Following compaction, the samples were shaped into cylinders with a diameter of 50 mm and a height of 100 mm, meeting the minimum compaction standard of 96% for high-grade highway subgrade material. Each sample was tightly wrapped in plastic to prevent moisture loss. According to the Test Methods of Materials Stabilized with Inorganic Binders for Highway Engineering (T0845) of the Chinese Specification (JTG E51-2009)41, samples were cured in a room maintained at 20 ± 2 °C and relative humidity above 95% for periods of 3, 7, 14, and 28 days. The sample preparation and curing processes are depicted in Fig. 3.

Fig. 3
figure 3

Samples molded and curing process.

Full size image

Table 3 outlines the blend design for the unconfined compressive strength test. Group 1 comprises fiber-reinforced silty sand, serving as a baseline for comparisons with Group 2, which contains fiber-reinforced cemented silty sand. Group 3, also consisting of fiber-reinforced cemented silty sand, includes various fiber lengths to facilitate a detailed comparative analysis with Group 2.

Table 3 The summary of test schemes.
Full size table

Test methods

Unconfined compressive strength test

The unconfined compressive strength test is primarily used to assess the efficacy of cement stabilization, drawing on insights from concrete testing practices. Following curing, the universal testing machine (Fig. 4) is utilized to perform the unconfined compressive strength test. The maximum load capacity of the sensor is 100 kN, with a measurement error of ± 1% and control precision of ± 0.5%. During testing, a constant vertical loading rate of 1 mm/min is applied to the sample, which is subjected to continuous loading until failure. To establish reliable data, at least three samples with identical properties are tested, the average of these three measurements is taken as the UCS value (\(q_{u}\)). The UCS (\(q_{u}\)) serves as a quantitative metric specifically designed to assess the performance of stabilized soil, as calculated using Eq. (1)42.

$$q_{u} = \frac{P}{{\pi R^{2} }}$$
(1)

where \(q_{u}\) is the value of UCS (MPa), P is the maximum vertical loading pressure (kN), and R is the radius of the cylinder (mm).

Fig. 4
figure 4

Universal testing machine and AE test system.

Full size image

AE test

Changes in AE parameters due to F–T cycles correlate with alterations in the soil's internal structure. The elastic wave, generated by internal structural damage, is transmitted to the AE sensor's surface, where it is converted into an electrical signal. A signal acquisition and processing system then calculates a set of signal parameters to characterize the digitized AE waveform.

The AE test is conducted concurrently with the unconfined compressive strength test to monitor the AE behavior of samples under uniaxial compression, as depicted in Fig. 4. The instrumentation includes a six-channel SAEU2S AE data acquisition system from Shenghuaxingye Technology Co., Ltd., Beijing, China. This system comprises a sensor, a preamplifier, data acquisition and processing equipment, a display system, and a connecting line.

Prior to testing, a SR150M 150 kHz resonant AE sensor with a frequency range of 60 to 400 kHz is affixed to the sample's side using duct tape. Vaseline is applied as a coupling agent to eliminate air at the contact interface. The preamplifier gain is set at 40 dB. Automatic exposure detection is configured with a threshold of 40 dB to exclude environmental noise, and low-pass and high-pass filters are adjusted to 20 kHz and 400 kHz, respectively, to eliminate electrical noise. The sampling frequency, peak definition time, hit definition time, and hit lock time are set at 5 MSPS, 50 µs, 150 µs, and 300 µs, respectively, ensuring the integrity and accuracy of signal acquisition43. A lead breaking test confirms proper AE sensor contact before sample loading and synchronous AE signal acquisition43,44.

AE testing typically yields parameters such as hits, ring count, energy, amplitude, rise time, and duration (Fig. 5). From these basic parameters, characteristic parameters like cumulative ringing count, cumulative energy, energy, amplitude, and RA and AF values, are derived. These parameters are essential for assessing the damage stages, crack development, and failure modes of BF-reinforced cemented silty soil under uniaxial compression under F–T cycles.

Fig. 5
figure 5

Schematic diagram of AE signal parameters.

Full size image

The cumulative ring count is the cumulative value of the number of signal exceeds the threshold, which characterizes the intensity and frequency of AE event activity, expressed in times, obtained by Eq. (2):

$$N = \sum\limits_{t = 0}^{{t = t_{i} }} {\frac{{f_{0} }}{\beta }\ln \frac{{V_{P} }}{V(t)}}$$
(2)

where: \(f_{0}\) is the response center frequency of the transducer,\(\beta\) is the attenuation coefficient of the wave, \(V_{P}\) is the peak voltage, and \(V(t)\) is the threshold voltage at the time t.

Energy is quantified as the area enclosed by the signal curve and the time axis, which reflects the intensity of AE event activity. It serves as a critical indicator for assessing material and structural damage. Energy is particularly sensitive to the detection of damage evolution, surpassing other basic parameters in its ability to provide detailed signal source information45. AE energy correlates with the relative strength and severity of cracks, generally, higher AE energy indicates stronger cracks, and lower energy suggests lesser severity46. AE energy and cumulative energy are quantified by Eqs. (3) and (4):

$$E_{AE} = \int\limits_{0}^{t} {V^{2} } (t)dt$$
(3)
$$CE_{AE} = \sum\limits_{t = 0}^{{t = t_{i} }} {E_{AE} } (t)$$
(4)

where \(E_{AE}\) and \(CE_{AE}\) represent the energy value and cumulative energy value, respectively, (mV·μs), and \(V(t)\) is the threshold voltage at the time t.

The amplitude is the maximum value of the corresponding event signal waveform, which characterizes the type of AE signal waveform and the intensity of AE event activity, which is represented by Eq. (5).

$$Amp = 20\lg \frac{{f(t)_{\max } }}{1\mu V} - G_{pre}$$
(5)

where Amp represents the amplitude (dB), \(f(t)_{\max }\) is the positive maximum value, when the sensing god outputs 1 μV is 0 dB,\(G_{pre}\) is the preamplifier gain.

RA and AF values are effective indicators for characterizing the crack mode of materials47. A high RA value coupled with a low AF value indicates that the failure mode of the material is predominantly shear. Conversely, a low RA value in conjunction with a high AF value suggests a tensile failure mode. RA and AF values are determined using Eqs. (6) and (7):

$$RA = \frac{Rise\; \, Time}{{Amplitude}}$$
(6)
$$AF = \frac{Counts}{{Duration}}$$
(7)

F–T cycle test

F–T cycles are more readily replicated in laboratory settings than in situ35. In Jilin Province's Songyuan County, where the silty sand is sourced, the average lowest temperature in January falls below − 20 °C, while the average highest temperature in July reaches approximately 28 °C. Consequently, the temperatures for the freezing and thawing phases of the cycles are set to − 20 °C and 20 °C, respectively, to reflect the actual geological conditions48. After a 14-day curing period, samples are placed in a digital refrigerator (Fig. 6) at − 20 °C for 24 h for freezing, followed by a 24-h thawing period at 20 °C.

Fig. 6
figure 6

The digital refrigerator for F–T cycles.

Full size image

Our previous research indicates that splitting tensile strength initially increases with fiber content, peaks at 0.2%, and then decreases. Additionally, splitting tensile strength increases with fiber length, with the most significant rise occurring after a 14-day curing period. Further curing time results in only marginal strength gains. Sections "Effect of fiber content" to "Effect of curing time" indicate that the maximum UCS is achieved at a fiber content of 0.2%, and UCS decreases with increased fiber length. The UCS shows the most significant increase after 14 days of curing, with only minimal increases as curing time continues. Therefore, considering the effects of fibers on splitting tensile strength and UCS, as well as engineering costs, the selected parameters for the freeze–thaw (F–T) cycle tests are 0.2% fiber content, 12 mm fiber length, and a 14-day curing period. Furthermore, numerical simulations suggest that the physical and mechanical properties of soil stabilize after 6 to 8 F–T cycles. Consequently, as shown in Table 4, the number of F–T cycles is set at 0, 1, 3, 6, and 10 cycles. After completing each set of F–T cycles, samples are removed for UCS and AE testing. The remaining samples continue F–T testing until all 10 cycles are completed.

Table 4 The summary of F–T cycles test schemes.
Full size table

Results and discussion

Unconfined compression strength

Effect of fiber content

Figure 7 displays the UCS of fiber-reinforced silty sand and fiber-reinforced cemented silty sand with varying fiber contents. As shown in Fig. 7a, without cement, the UCS increases as the fiber content rises from 0.0% to 0.1% and then to 0.2%, with corresponding values of 0.45 MPa, 0.48 MPa, and 0.67 MPa. However, when the fiber content further increases to 0.3% and 0.4%, the UCS decreases to 0.46 MPa and 0.49 MPa, respectively. This trend suggests that the UCS initially increases and then decreases with increasing fiber content, peaking at 0.2%. The decline in strength beyond 0.2% fiber content is attributed to the agglomeration effect of the fibers, which adversely affects the material's strength. Consequently, a 0.2% fiber dosage offers the optimal enhancement of strength when only fibers are included. Similarly, Fig. 7b illustrates that the UCS variation in fiber-reinforced cemented silty sand follows a comparable pattern, also achieving maximum strength at a 0.2% fiber dosage.

Fig. 7
figure 7

Unconfined compression strength at varying fiber content: (a) fiber reinforced silty sand and (b) fiber reinforced cemented silty sand.

Full size image

Effect of fiber length

Figure 8 demonstrates that fiber length significantly impacts the UCS of SM2P0.2B, showing a decline in UCS as fiber length increases at different curing times. The greatest increase in UCS, 1.32 times that of cemented silty sand, occurs with a fiber length of 6 mm and a curing time of 28 days, followed by 1.17 times at 12 mm, and the minimum increase, 1.01 times, at 18 mm. This pattern suggests that shorter fiber lengths facilitate a more uniform distribution among particles, enhancing pore-filling and thus increasing strength.

Fig. 8
figure 8

Unconfined compression strength of SM2P0.2B at varying fiber lengths of different curing time.

Full size image

When the fiber length is 6 mm and the curing times are 3, 7, 14, and 28 days, the respective UCS values are 1.48 MPa, 1.74 MPa, 1.88 MPa, and 2.08 MPa. For a fiber length of 12 mm, the strength decreases by 5.68%, 15.82%, 11.41%, and 12.17% respectively, compared to 6 mm. With fibers of 18 mm, the strength further decreases by 10.19%, 5.21%, 13.33%, and 15.26% respectively, compared to 12 mm. These results indicate that the decrement in strength amplifies with increased fiber length and stabilizes after a curing period of 14 days.

Effect of curing time

Figure 9 illustrates that the UCS of both fiber-reinforced silty sand and fiber-reinforced cemented silty sand increases with curing time, reaching maximum values after 28 days. Specifically, the UCS of SM0P0.2B is recorded at 0.32 MPa, 0.56 MPa, 0.60 MPa, and 0.67 MPa at 3, 7, 14, and 28 days, respectively, with increments of strength between adjacent curing periods being 75.00%, 7.14%, and 11.67%. The most significant increase occurs between the 3-day and 7-day marks. For SM2P0.2B, the UCS is 1.39 MPa after 3 days of curing, with subsequent increases of 7.16%, 13.22%, and 10.00% for each curing period. The inclusion of cement shows that the most substantial strength enhancement occurs at 14 days, which can be attributed to the cement hydration reaction. This suggests that a relatively complete hydration reaction has taken place by this point, significantly improving the material's strength.

Fig. 9
figure 9

Unconfined compression strength with different curing time: (a) fiber reinforced silty sand and (b) fiber reinforced cemented silty sand.

Full size image

Effect of F–T cycles

The change of UCS

Figure 10 demonstrates that the UCS of SM0P0B, SM0P0.2B, SM2P0B, and SM2P0.2B decreases with the increase in the number of F–T cycles after a curing period of 14 days. Initially, without F–T cycles, the strengths of SM0P0B, SM0P0.2B, SM2P0B, and SM2P0.2B are 0.41 MPa, 0.60 MPa, 1.24 MPa, and 1.68 MPa, respectively. This data implies that the incorporation of single-doped fiber, single-doped cement, and a comprehensive fiber-cement modification enhances strength by factors of 0.45, 2.02, and 3.10, respectively. Throughout 1 to 10 F–T cycles, the strength increase amplitude for SM0P0.2B was 0.17, 0.37, 0.76, and 0.33 times, respectively, under various conditions. For SM2P0B, these increases were more pronounced, at 2.54, 4.44, 4.12, and 6.59 times, respectively. For SM2P0.2B, the increments were even greater, at 4.52, 8.09, 8.98, and 10.76 times, respectively. These findings indicate that the amplitude of strength increase for both cemented and fiber-reinforced cemented soil stabilizes after three cycles. The order of effectiveness in enhancing strength under F–T cycles is: fiber-cement modification > single-doped cement > single-doped fiber, with comprehensive fiber-cement modification showing the highest resistance to F–T cycles.

Fig. 10
figure 10

Unconfined compression strength at 14 days of curing under F–T cycles: (a) SM0P0B, (b) SM0P0.2B, (c) SM2P0B and (d) SM2P0.2B.

Full size image

The loss of UCS

The UCS loss of SM0P0B, SM0P0.2B, SM2P0B and SM2P0.2B under F–T cycles can be obtained by Eq. (8).

$$q_{{{\text{ul}}}} = \left( {1 - \frac{{q_{{{\text{u }}N_{ft} = n}} }}{{q_{{{\text{u }}N_{ft} = 0}} }}} \right) \times 100\% ,n = 1,3,6,10$$
(8)

where \(q_{{{\text{ul}}}}\) is the strength loss percentage (%), \(q_{{{\text{u }}N_{ft} = n}}\) is the UCS at n F–T cycles (MPa) and \(q_{{{\text{u }}N_{ft} = 0}}\) is the UCS at 0 F–T cycles (MPa).

Figure 11 demonstrates that the strength loss of samples under a fixed number of F–T cycles follows an increasing trend: SM2P0.2B < SM2P0B < SM0P0B < SM0P0.2B. The data indicate that the addition of single-doped fibers reduces the F–T stability of silty sand. The impact of three cycles on strength is particularly pronounced, with strength reductions of 74.20% for SM0P0B, 75.71% for SM0P0.2B, 53.51% for SM2P0B, and 42.76% for SM2P0.2B. After 6 to 10 cycles, the strength loss of SM0P0B, SM0P0.2B, SM2P0B, and SM2P0.2B stabilizes, ranging from 69.55 to 76.83%, 63.13 to 78.79%, 41.86 to 48.35%, and 25.87 to 33.65%, respectively. These values suggest that average F–T strength losses are approximately 73% for SM0P0B, 75% for SM0P0.2B, 45% for SM2P0B, and 30% for SM2P0.2B. Incorporating fibers into cemented silty sand significantly enhances its F–T stability. This enhancement is attributed to the F–T cycles promoting water migration within the silty sand, which substantially alters particle connections and arrangements through repeated freeze–thaw actions. The inclusion of only fibers results in a significant pull-out effect under mechanical loads and F–T conditions, leading to looser particle bonds. In contrast, the addition of cement substantially improves particle adhesion, as the hydration reactions form a cement-soil skeleton that enhances both the integrity and F–T stability of the silty sand. When fibers and cement are combined, this structure effectively prevents fiber pull-out, enhances the bridging effect of the fibers, and results in a tighter particle connection, thereby yielding optimal F–T stability.

Fig. 11
figure 11

The loss of unconfined compression strength under F–T cycles.

Full size image

Analysis of AE characteristic under F–T cycles

Cumulative ring count and cumulative energy

Damage to fiber-reinforced cemented silty sand under F–T cycles can be categorized into three stages: micro-crack initiation, micro-crack propagation, and macro-crack penetration, as indicated by cumulative ring count and cumulative energy (Fig. 12).

Fig. 12
figure 12

Curve of cumulative count and cumulative energy with time under F–T cycles: (a) Nft = 0, (b) Nft = 1, (c) Nft = 3, (d) Nft = 6 and (e) Nft = 10.

Full size image

Initially, after 6 F–T cycles, the micro-crack initiation phase is notably brief (Fig. 12d), indicating an increased formation of micro-cracks within the sample. Comparing Fig. 12b-e with Fig. 12a shows that the rates of increase in cumulative ring count and cumulative energy are significantly decelerated after F–T cycles, suggesting that these cycles induce early local damage and slow the propagation of micro-cracks. The cumulative ring counts of 1500, 1000, 400, 200, and 600 correspond to cumulative energy values of 1800 mv μs, 1200 mv μs, 600 mv μs, 300 mv μs, and 600 mv μs, respectively, illustrating that samples after 0, 1, 3, 6, and 10 cycles reach critical failure points. These findings indicate that F–T cycles significantly reduce both cumulative ring count and cumulative energy at these points, while also diminishing the frequency and intensity of AE activity, which suggests early local damage and decreased load-bearing capacity. After 6 cycles, the cumulative ring count and cumulative energy at the critical failure point are at their minimum, indicating that this number of cycles causes the most severe internal damage to the sample and represents the least favorable bearing state. In the final stage, the curvature of the cumulative ring count and cumulative energy curves increases significantly compared to the previous stage, indicating rapid propagation of macro-cracks and culminating in sample failure.

Energy

Figure 13 demonstrates that the damage of the sample under F–T cycles can be categorized into three stages based on energy, aligning with divisions made according to cumulative ring count and cumulative energy.

Fig. 13
figure 13

Curve of AE energy and load with time under F–T cycles: (a) Nft = 0, (b) Nft = 1, (c) Nft = 3, (d) Nft = 6 and (e) Nft = 10.

Full size image

Figure 13a-e illustrate that the initial propagation load for micro-cracks is 0.3 kN for 0 and 1 F–T cycles, decreasing to approximately 0.2 kN at 3, 6, and 10 cycles. This reduction demonstrates a decline in the bearing capacity of the sample during the micro-crack propagation stage, which stabilizes after 3 cycles. The critical failure loads at various stages are 3.0 kN, 1.5 kN, 1.3 kN, 1.8 kN, and 1.6 kN, respectively. This pattern indicates that critical failure loads decrease as the number of F–T cycles increases, with the most notable reduction occurring after the first cycle. Figure 13d,e reveal that after 6 and 10 cycles, the excitation energy during the micro-crack propagation stage is minimal, and the acoustic emission (AE) energy signal intensity is low. This suggests a high occurrence of micro-cracks within the sample, signifying extensive F–T damage at this stage, which corresponds to the most adverse bearing condition. Consequently, the mechanical properties observed after 6 cycles can serve as a reference for construction practices.

Amplitude

Figure 14 illustrates that the damage of samples subjected to F–T cycles can be divided into three stages based on amplitude, correlating closely with the damage stages of the sample. Figure 14a demonstrates that in the micro-crack propagation stage, samples undergoing F–T cycles produce a high number of amplitude peaks, indicative of active AE amplitude signals and rapid micro-crack propagation, with the amplitude at the critical failure point reaching 65 dB. Figure 14b-e show that the number of amplitude peaks in the micro-crack propagation stage decreases with successive F–T cycles, confirming that these cycles induce micro-cracks within the sample. Notably, by the sixth cycle (Fig. 14d), the amplitude is at its lowest and the AE signal intensity is the weakest, indicating significant F–T damage due to the proliferation of micro-cracks. The boundary amplitudes at the critical failure points for 1, 3, 6, and 10 cycles are recorded at 75 dB, 68 dB, 67 dB, and 67 dB, respectively. This trend suggests that the amplitude at the critical failure point increases initially and then stabilizes after six cycles, serving as a potential early warning indicator of sample damage under F–T conditions. In the third stage, the amplitude of samples not subjected to freezing and thawing decreases, suggesting that macro-cracks have reached a stable propagation phase. Conversely, after multiple F–T cycles, the amplitude increases significantly, indicating rapid macro-crack propagation on the foundation of micro-cracks exacerbated by the cycles and additional loading, thereby accelerating sample destruction.

Fig. 14
figure 14

Curve of amplitude and load with time under F–T cycles: (a) Nft = 0, (b) Nft = 1, (c) Nft = 3, (d) Nft = 6 and (e) Nft = 10.

Full size image

RA and AF

To effectively illustrate the trends of AE parameters, each data point on the curve represents the moving average of the last 20 events. Figure 15 demonstrates that the RA and AF curves display continuous fluctuations over time due to the combined effects of shear and tensile forces during the failure process. Without F–T cycles, the AF value trends downward, while the RA value exhibits stable fluctuations across three stages, indicating that shear failure predominates. During F–T cycles, the RA and AF curves show consistent fluctuations in the first two stages. In the third stage, after 1 and 3 cycles, a significant increase in RA and a marked decrease in AF are observed, suggesting that shear failure dominates during the macro-crack propagation stage. After 6 and 10 cycles, the RA value decreases and the AF value increases, indicating a shift towards tensile failure as the predominant mode. This shift from shear to tensile failure after multiple F–T cycles is attributed to internal water migration caused by F–T, leading to particle collapse. The enhanced bridging effect of fibers after several F–T cycles becomes apparent, ultimately resulting in a transformation in the failure mode under load.

Fig. 15
figure 15

Curve of RA and AF with time under F–T cycles: (a) Nft = 0, (b) Nft = 1, (c) Nft = 3, (d) Nft = 6 and (e) Nft = 10.

Full size image

Conclusions

In this study, the UCS of fiber-reinforced cemented silty sand was assessed under various conditions, including alterations in fiber content, fiber length, curing time, and F–T cycles. The damage process during F–T cycles under uniaxial compression was monitored using AE technology. The experimental data led to the following conclusions:

  1. 1.

    The UCS of fiber-reinforced cemented silty sand initially increased and then decreased with fiber content, reaching a peak at an optimal fiber content of 0.2%. UCS increased with curing time but decreased with longer fiber lengths and more F–T cycles.

  2. 2.

    After stabilization, the F–T strength loss for silty sand, fiber-reinforced silty sand, cemented silty sand, and fiber-reinforced cemented silty sand was approximately 73%, 75%, 45%, and 30%, respectively. The combination of fibers and cement provided the greatest F–T stability for silty sand.

  3. 3.

    Damage stages during F–T cycles were effectively categorized into three distinct phases based on the AE characteristic parameters of the fiber-reinforced cemented silty sand. F–T cycles tended to reduce the frequency and intensity of AE signals, leading to early localized damage.

  4. 4.

    By the sixth F–T cycle, the excitation energy was minimal, indicating severe internal damage. Therefore, the mechanical properties observed after six cycles could serve as a reference for construction practices.

  5. 5.

    A significant change in AE amplitude, notably reaching 67 dB, marked a transition in damage stages and could serve as an early warning signal for potential failure of the fiber-reinforced cemented silty sand subgrade.