1. Introduction
Tunnel engineering in red-layer soft rock represents one of the significant challenges in transportation infrastructure construction. As shown in Figure 1, such rock masses are widely distributed across northern Europe; southwestern and northwestern China; and extensive regions including the western United States and Africa. Their unique engineering geological characteristics make tunnel construction and support design highly complex and risky technical challenges [1]. Red-layer soft rock primarily consists of Neogene argillaceous sandstone and sandy conglomerate, characterized by poor diagenesis, weak cementation, and water-induced softening. It falls within the category of typical extremely soft to soft rock [2]. During tunnel excavation, this rock mass is highly susceptible to stress redistribution due to excavation disturbance, leading to engineering hazards such as significant surrounding rock deformation, support structure failure, and even tunnel collapse [3]. Laboratory tests and field monitoring indicate that the long-term creep rate of red-bed soft rock ranges from 0.02 to 2 mm/day, with stabilization typically requiring 90 to 120 days or more after excavation. This prolonged rheological behavior leads to cumulative deformation exceeding 70 mm in unsupported sections, significantly impacting tunnel stability. The rapid deformation and persistent rheological behavior of surrounding rock during construction impose stringent demands on support systems [4]. Conventional support designs often prove inadequate for effectively controlling rock stability. There is an urgent need to systematically investigate the dynamic interaction mechanisms between surrounding rock and support structures in red-layer soft rock tunnels to provide a scientific basis for engineering practice.
The design and optimization of tunnel support systems constitute the core issue for ensuring engineering safety [5,6,7,8,9,10,11]. The primary support system (e.g., shotcrete and steel arches) must provide sufficient stiffness within a short duration to restrain the deformation acceleration phase, while the secondary lining assumes long-term stabilization functions. Monitoring data indicate that the typical deformation curve of soft rock tunnels comprises three stages [12]: a rapid deformation period (1–15 days post-excavation); a gradual adjustment period (15–90 days); and a stable convergence period (beyond 90 days). As an active support measure, the mechanical behavior of rock bolts directly influences the development and control of the surrounding rock loosened zone (a rock mass zone around excavations where stress redistribution causes fracturing and reduced integrity, directly affecting support design). However, excessive reliance should be avoided, particularly in soft rock tunnels [13]. Meanwhile, the synergistic effects between steel arches and shotcrete in primary support determine the capacity to restrain surrounding rock deformation [14]. The secondary lining, serving as the ultimate load-bearing structure, has its installation timing and stress characteristics critically impacting long-term tunnel stability [15].
Regarding rock mass characteristics, scholars have revealed through laboratory tests that the softening coefficient of red-layer soft rock is as low as 0.3–0.5, with uniaxial compressive strength generally below 15 MPa. Its long-term creep rate can reach 0.1–0.3 mm per day. These properties result in an extremely weak self-stabilization capacity of surrounding rock, making large-scale loosened zones highly prone to form after excavation [16]. At the support theory level, flexible support systems based on the New Austrian Tunneling Method (NATM) concept of “surrounding rock-support interaction” are widely applied. This approach emphasizes providing progressive support resistance while allowing moderate deformation of surrounding rock through combined support forms including rock bolts, shotcrete, and steel arches [17]. For instance: rock bolts improve the stress field of surrounding rock through axial tension, with their anchoring effect significantly enhancing the cohesion and internal friction angle of the rock mass; shotcrete achieves transformation of three-dimensional stress states by sealing rock surfaces and filling steel frame gaps, thereby restraining rock relaxation; and the high stiffness of steel arches enables rapid control of initial deformation, playing a critical supporting role, especially during bench cut excavation. In monitoring technology, conventional methods typically employ instruments such as displacement meters and pressure cells [18], achieving multi-angle monitoring of surrounding rock pressure, support internal forces, and deformations.
Current research predominantly focuses on conventional red-layer soft rock, with limited investigation into Neogene red-layer soft rock. There exists a scarcity of data concerning the stress behavior of support structures in Neogene red-layer soft rock tunnels under traditional support modes. Due to its young geological age, poor diagenesis, and weak rock properties, Neogene red-layer formations exhibit low strength and deformation modulus, significant deformation, susceptibility to water-induced softening, and pronounced rheological behavior [19,20,21], while tunnel excavation disturbances further degrade surrounding rock properties. Moreover, existing support designs heavily rely on empirical formulas or code-recommended values [22], lacking targeted consideration for the heterogeneity, anisotropy, and environmental sensitivity of red-layer soft rock. This study investigates and analyzes the stress deformation mechanisms and real-time performance of tunnel support in Neogene red-bed soft rock. Using wireless remote monitoring, it quantifies a 2.5 m thick loose zone, reveals stress asymmetry resulting from rock heterogeneity and rheological properties, identifies critical support timings, and enables data-driven optimization of support for such strata. For instance, rock bolt length design generally adopts the empirical rule of “loosened zone thickness plus safety margin”, yet determining loosened zone thickness still depends on indirect methods such as core drilling or elastic wave testing, whereas the strong rheology of red-layer soft rock makes dynamic expansion of the loosened zone difficult to predict accurately. Steel arch stiffness selection faces the dilemma between “over-constraint” and “inadequate support”: while rigid steel sections can rapidly suppress deformation, they risk structural failure due to concentrated rock pressure, whereas flexible lattice girders facilitate stress release but struggle to cope with sudden deformations [23,24,25,26]. Secondly, the application of monitoring data remains at the single-parameter analysis level [27], with no established framework for the fusion modeling of multi-source heterogeneous data (e.g., surrounding rock pressure, support internal forces, temperature–humidity, and deformation rates) and real-time feedback mechanisms, resulting in support efficiency assessments lagging behind engineering demands. Furthermore, quantitative standards for support timing selection are lacking [28]; current codes specify “stabilized surrounding rock deformation” as the condition for secondary lining construction, yet the rheological characteristics of red-layer soft rock make deformation stability thresholds difficult to define, often leading to extreme cases of premature load-bearing by secondary lining or primary support failure in practice. This study implements dynamic monitoring of surrounding rock contact pressure, internal force distribution in support structures, rock mass-lining cooperative deformation characteristics, and temperature–humidity parameters at the Huizhou No. 1 and No. 2 Neogene red-layer tunnels through integrated remote online monitoring and construction site monitoring systems. Based on multi-source monitoring data, this study systematically clarifies the load transfer paths and spatiotemporal evolution laws of the support system for Neogene red-bed tunnels and then proposes a four-step optimization design method: multi-parameter monitoring, spatiotemporal mechanism diagnosis, dynamic feedback of support parameters, and collaborative control of construction sequence. This method achieves an engineering adaptability iteration through wireless remote monitoring, providing a universal design paradigm for similar weak rock tunnels.
2. Materials and Methods
2.1. Test Subjects
The Huizhou Tunnel is situated at the transitional margin of the Hui-Cheng Basin, geomorphologically belonging to a low–middle mountainous hilly area formed by denudation. The groundwater system in the tunnel area primarily exists as fracture water in bedrock, with atmospheric precipitation infiltration serving as its main recharge source. The water content of the Neogene ranges from 1.38% to 18.08%, with an average of 9.25%. Specifically, it is 12.35% for mudstone, 12.28% for sandstone, and 7.12% for glutenite. The average water content of the Neogene is 10.94%. The water absorption ranges from 0.56% to 18.12%, with an average of 7.46%. Specifically, the maximum value for mudstone is 10.79%, while the minimum value for bottom conglomerate is 1.44%. The water absorption of sandstone is 6.82%, and that of glutenite is 6.98%. The regional strata consist of Neogene argillaceous sandstone and sandy conglomerate (N2): reddish-brown in color, weakly cemented by argillaceous material, exhibiting poor diagenesis and medium-thick layered structure. The maximum particle size observed in sandy conglomerate reaches 3–4 cm, with grains predominantly subrounded. This formation shows susceptibility to water-induced softening and is classified as extremely soft to soft rock, as shown in Figure 2.
Furthermore, to evaluate the rock mass quality, borehole acoustic logging and rock block wave velocity tests were conducted in exploration boreholes within the tunnel site area. These assessments of rock strength and integrity, performed in accordance with the relevant requirements of the “Specifications for Highway Engineering Geological Investigation” [29], also served as a reference for lithological stratification. The results indicate that strongly weathered and moderately weathered sand conglomerate and argillaceous sandstone, with integrity coefficients (Kv) ranging from 0.21 to 0.54, are mostly classified as relatively fragmented to fragmented. These materials exhibit low strength and are prone to softening and disintegration. In contrast, slightly weathered rock masses (Kv = 0.61~0.82) show relatively good integrity, yet they account for a limited proportion of the total. Detailed results are shown in Table 1. Consequently, focused prevention measures are required against the risks of differential deformation and local collapse at the interfaces between soft and hard interbedded layers.
Based on the above analysis, the stratigraphic lithology and engineering geological conditions at the Huizhou Tunnel site fully demonstrate the typical attributes of Neogene red-bed soft rock: lithologically characterized by an interbedded structure of weakly cemented argillaceous sandstone and sandy conglomerate, mechanically exhibiting core deficiencies of low strength and susceptibility to softening/disintegration; structurally featuring significant heterogeneity with alternating soft-hard interlayers, compounded by gently dipping bedding planes and differential weathering effects that form structurally weak interfaces; and hydrogeologically displaying prominent dynamic sensitivity in the red-bed fissure water system, where permeability–softening coupling intensifies engineering complexity. This series of characteristics—poor diagenesis, sensitive hydro-physical properties, and strong structural heterogeneity—collectively constitute the quintessential engineering geological system of Neogene red-bed soft rock formations. Given the typicality and uniqueness of Huizhou Tunnel’s geological conditions, it is necessary to establish it as a paradigmatic case study for the systematic analysis of Neogene red-bed soft rock tunnels. Through in-depth research on its geo-mechanical coupling mechanisms and construction response characteristics, a theoretical–practical paradigm of “special geology identification → surrounding rock deterioration control → structural synergistic bearing (collaborative load resistance between support structures and surrounding rock, where supports restrain deformation and rock provides reactive forces)” can be developed, which will not only provide refined construction references for tunnel projects with analogous lithological assemblages (weakly cemented soft rock interbeds + active fissure water) but also substantively support the advancement of life-cycle disaster prevention systems for red-bed soft rock tunnels.
2.2. Infinite Remote Monitoring System
The structural health monitoring (SHM) of tunnels [30] refers to the long-term online monitoring of various mechanical responses of tunnel structures during operation by installing sensors at key locations, enabling real-time early warning and stability assessment to ensure safe tunnel operation. Data are collected manually. Given the relatively long monitoring period, they are gathered at 24 h intervals. A comparison of advantages and disadvantages between online structural health monitoring and traditional manual inspections is presented in Table 2.
Based on the engineering and hydrogeological characteristics of red-bed soft rock, combined with red-bed rock mass survey results and tunnel construction progress, remote wireless monitoring systems were deployed at typical Neogene red-bed argillaceous sandstone sections in Huizhou Tunnel No. 2 and Huizhou Tunnel No. 1 along the Liangdang to Huixian Expressway to conduct monitoring measurements of tunnel surrounding rock pressure, internal forces in support structures, deformations of rock mass and support measures, and rock temperature–humidity conditions. The selection of monitoring points in this study is comprehensively determined based on geological conditions, design features, construction risks, and monitoring objectives: it focuses on geological weak zones (such as faults and water-rich strata), key structural components (arch crown, side walls, invert, etc.), and high-risk areas like large cross-sections or shallow buried sections. Meanwhile, in conjunction with construction methods and special working conditions, it ensures that monitoring points cover deformation and stress concentration areas to validate the design, guide construction, warn of risks, and assess long-term safety. See also the study by Xie Quanmin et al. [31]. The monitored parameters, instrumentation, and number of measurement points are detailed in Table 3, with the sensor layout illustrated in Figure 3.
3. Results
3.1. Mechanical Behavior and Deformation Characteristics of Anchor Bolts
To quantify the variability of monitoring data, this study performed moving average filtering on the time-series data of bolt strain (Figure 4) to eliminate instantaneous construction interferences. The significant correlation between bolt stress and surrounding rock temperature/humidity was verified using the Pearson correlation coefficient (|r| > 0.85, p < 0.01). To reduce measurement uncertainties, all sensors were calibrated in the laboratory with a systematic error ≤±0.5%FS. The section layout adopted a triple redundant arrangement, with three groups of symmetric measuring points set for each monitoring section, and outliers with deviations >15% were excluded.
As shown in Figure 4, the time-dependent stress–strain curves of monitoring section rock bolts reveal high sensitivity to surrounding rock pressure after installation, exhibiting three-phase characteristics: rapid deformation, gradual deformation, and stabilization. Arch crown bolts experience maximum stress within the rock loosening zone, while the peak stress shifts to the haunch area in transitional rock segments, demonstrating significant asymmetry and anisotropy, as illustrated in Figure 5. Construction disturbances triggered V-shaped/stepped abnormal stress redistribution; although transient, this resulted in widespread axial force increases post-stabilization, with most bolts effectively participating in rock arch formation. An anomalous stress fluctuation detected at the left haunch bolt (1.5 m depth) retains collapse warning significance. The specific potential causes are as follows. Stress concentration induced by asymmetric loading: Due to heterogeneous rock properties and stress redistribution caused by layered excavation, the left haunch resides in a high-stress zone. Thermomechanical coupling effect: The temperature surge to 147 °C most likely stems from frictional heating at the bolt–grout interface under cyclic loading, validated by the correlation between stress peaks and thermal anomalies. Loosened zone expansion: The maximum stress at 2.5 m depth aligns with the thickness of the loosened zone, indicating that the left haunch bolts are partially anchored in loosened rock mass, weakening the anchorage effect. In the Neogene argillaceous sandstone tunnel, rock bolt axial forces exhibit large variation amplitudes and prolonged stabilization periods, with peak values occurring at 2.5 m depth—consistent with rock loosening zone thickness determined by P-wave velocity/seismic tomography tests, thereby providing critical references for bolt parameter design (effective length and external diameter) and support efficacy evaluation.
Figure 6 presents the stress diagrams of primary lining rebar stress meters at the outer, inner, and middle positions on rock bolts in the YK35 + 470 section of Huizhou Tunnel No. 1; subplots (a), (b), and (c), respectively, display stress variation curves for rebar stress meters at the outer, middle, and inner measurement points on bolts installed in the right haunch, left haunch, right spandrel, left spandrel, and arch crown positions of the primary lining during monitoring.
As shown in Figure 6, the compressive stress at the inner ends of primary lining rock bolts ranges from 105 to 330.5 MPa, with an average of 233.2 MPa; the mid-section stress varies between 67 and 305.3 MPa, averaging 215.58 MPa; and the outer stress measures 140.9–276.5 MPa, with a mean value of 209.2 MPa. The close proximity of average stress values across the inner, middle, and outer sections demonstrates effective pressure-bearing capacity throughout the bolt length. Minimum stress values for all three sections consistently occur at the right spandrel, while peak stresses at the mid and inner sections are recorded at the right haunch. Overall, significantly elevated stresses at the arch crown and bilateral haunches indicate substantial deformation under high pressure. The cross-sectional stress distribution reveals symmetrical patterns across all measurement positions, consistently showing higher stresses at the crown and haunches but lower values at spandrels, confirming balanced bilateral loading symmetry. Throughout the monitoring period, bolt stresses remained stable, with negligible fluctuations at all locations except for anomalous variations observed at the inner and outer edges of the left haunch bolts.
3.2. Mechanical Behavior and Deformation Characteristics of Primary Lining
Figure 7 and Figure 8 present the time-dependent stress–strain curves of primary lining steel arches. Monitoring data indicates that the arches exhibit high sensitivity to constraining surrounding rock pressure: compressive stresses surge rapidly following upper bench excavation before stabilizing after lower bench excavation. Steel arches at sidewalls and wall footings experience relatively lower forces with alternating tension–compression states, while invert closure stabilizes stresses across all measurement points, demonstrating that prompt structural closure optimizes load distribution. Significantly higher stresses consistently occur at the arch crown and haunches (particularly near the central partition wall), whereas sidewalls and footings sustain lower but dynamically alternating loads. In common with previous studies, the stress–strain behavior of steel arches exhibits high sensitivity to surrounding rock pressure, with stresses rising rapidly after excavation and stabilizing with construction progress, and significant stress concentration exists in key positions such as the arch crown and arch waist. The difference is that Figure 7 and Figure 8 reveal the asymmetric characteristics of steel arch stress, such as the stress on the right arch shoulder being significantly higher than that on the left, reflecting the unique stress-bearing laws of red-bed soft rock under specific geological conditions. Crucially, invert closure substantially reduces structural stresses, confirming that the timely formation of closed-loop structures enhances stress-state management and deformation control.
Figure 9 and Figure 10 demonstrate that the development of shotcrete support forces progresses through four distinct phases, slow growth, rapid growth, decelerated growth, and stabilization, with phase duration ratios varying according to concrete curing time and steel arch type. Within the zone of spatial effects from the excavation face, support stiffness exhibits significant spatial variation, with stiffness growth rate decreasing as distance from the initial support point increases. During early-stage support application, concrete stiffness remains low with limited load-bearing capacity. Earlier attainment of design strength in shotcrete correlates with superior control of surrounding rock displacement. Furthermore, lattice girders exhibit higher sensitivity to shotcrete’s early-strength characteristics than steel sets.
The primary support system comprising steel arches and shotcrete effectively controls surrounding rock deformation and plastic zone (an area in rock where stresses exceed yield strength, inducing permanent deformation and influencing underground stability) radius development, with deformation rates significantly decreasing post-application. When support installation timing is optimized, the rock mass may not develop softening or residual zones while maintaining structural integrity, demonstrating that appropriately designed and timely installed primary support sufficiently ensures tunnel stability—rendering secondary lining primarily a safety reserve. Generally, the ground characteristic curve has a minimum point, corresponding to the minimum support reaction force, which represents the optimal support timing. Therefore, the key to rational support timing design for soft rock tunnels is to adjust the reserved deformation so that the support characteristic curve intersects the minimum point of the ground characteristic curve. Monitoring shows that the optimal support timing is within 4 h after excavation. Steel sets outperform lattice girders in controlling rock displacement and plastic zone expansion, achieving stabilization faster but bearing higher support loads, indicating that lattice girders’ lower initial stiffness facilitates beneficial stress redistribution. Crucially, earlier arch installation yields more pronounced deformation control yet proportionally increases imposed support forces on the rock mass. A detailed comparison is shown in Table 4.
As shown in Figure 11, temperature fluctuations at all embedded points exhibited significant initial variations before predominantly stabilizing toward the monitoring period’s conclusion. The arch crown stabilized at approximately 18 °C, the left haunch at 20 °C, the right haunch at 17 °C, the left spandrel between 17 and 20 °C, and the right spandrel at 18 °C, indicating normalized thermal conditions and confirming stable rock mass conditions. During initial monitoring, the arch crown registered the highest temperature (26.5 °C at the external rebar meter), while mid-March fluctuations in central/internal zones correlated with ambient temperature or heat transfer through soft rock layers. The left haunch’s internal sensor recorded dramatic fluctuations (−5 °C to 147 °C), directly corresponding to stress variations in rock bolts, whereas external/mid-point temperatures remained constant at 19.6 °C; the correlation is shown in Figure 12. The right haunch’s central measurement point showed considerable thermal volatility, consistent with its status as the highest-stress zone, demonstrating temperature’s significant influence on soft rock mechanics. Both spandrels exhibited similar thermal patterns despite minor extremum differences. Crucially, temperature data correlates convincingly with rebar stress meter readings, validating monitoring reliability and establishing critical benchmarks for future structural assessments.
3.3. Mechanical Behavior and Deformation Characteristics of Secondary Lining
Figure 13, Figure 14 and Figure 15 illustrate the development of contact pressure between the initial support and the secondary lining throughout the entire process, from the completion of secondary lining concrete pouring, through the gradual increase in concrete strength, to the stress redistribution stabilizing after formwork removal. The process of load transfer from the primary support to the secondary lining can be divided into three stages: Before 15 March in Figure 12: After concrete casting, when the secondary lining has not yet developed strength, the primary support bears all loads, and the contact pressure remains stable. 15–27 March in Figure 12: During the strength growth period, the development of concrete strength increases the lining stiffness, causing gradual load transfer to the secondary lining. The contact pressure increases rapidly. Adjustment and stabilization period: Due to the coupling effect of the surrounding rock rheological effect and lining stiffness, the contact pressure decreases temporarily and then slowly recovers, finally stabilizing approximately 20 days after formwork removal.
For the upper monitoring points on the arch walls, the contact pressure between the initial support and secondary lining undergoes an initial increase, followed by a decrease, and then a slow increase before gradually stabilizing. The lower section exhibits a delayed response, as the transfer of surrounding rock pressure from the initial lining to the secondary lining requires time, after which it undergoes a similar change process to the upper section. After the secondary lining concrete is poured, the contact pressure between the initial support and secondary lining continuously increases with the rapid gain in concrete strength and stiffness. Due to the “reaction force” provided by the formwork trolley, the contact pressure peaks during formwork removal, representing the most critical stress state for the secondary lining. To ensure the safety of the secondary lining, formwork removal should be delayed as much as possible to guarantee a gradual reduction in the formwork trolley’s ‘’reaction force’’. After formwork removal, the secondary lining transitions from a triaxial stress state to a biaxial stress state. As the “reaction force” from the trolley is released, the stress state between the surrounding rock and support continuously adjusts, leading to a gradual decrease in contact pressure between the initial support and secondary lining. However, due to the inherent rheological properties of the rock mass and the slow increase in stiffness of the cast-in-place secondary lining concrete after formwork removal, the contact pressure between the initial support and secondary lining slowly rises again. Generally, the contact pressure between the initial support and secondary lining stabilizes approximately 20 days after secondary lining formwork removal.
Based on the above, the contact pressure between the initial support and secondary lining reaches its maximum during formwork removal, at which point the secondary lining structure is under its most unsafe stress state. The load actually borne by the secondary lining is the deformation pressure of the surrounding rock transmitted through the initial support, which differs from the loosening pressure calculated using the loose body height specified in design codes. An analysis of the secondary lining’s safety under both code-specified loads and measured loads across different surrounding rock grades reveals that under code loads, the most unfavorable position of the tunnel structure is typically at the crown. In contrast, under measured loads, the critical section often occurs at localized stress concentration points. For this specific tunnel, the most critical location is at the arch shoulder. Furthermore, the secondary lining exhibits significantly higher safety margins under measured loads compared to those under code-specified loads. The specific safety margin is calculated according to Section 9.2 of JTG 3370.1-2018 [32], Design Specifications for Highway Tunnels, Volume 1, Civil Engineering. The formula is as follows:
(1)
where R denotes the structural resistance determined by material strength and sectional dimensions and S represents the load effect derived from both code-specified loads and measured loads.For the monitoring points at the tunnel invert, the variation pattern of contact pressure between the initial support and secondary lining is more complex. As shown in Figure 16 and Figure 17, upon completion of the secondary lining invert concrete pouring, the contact pressure at the invert and arch springing monitoring points measured 40 kPa and 25 kPa, respectively, essentially equivalent to the compressive stress induced by the self-weight of the secondary lining invert. However, with subsequent construction activities, the contact pressure at the invert position continuously adjusted. At the invert monitoring point, the contact pressure generally exhibited an initial increase, followed by a decrease before stabilizing, peaking at 75 kPa 7 days after invert pouring and eventually stabilizing at 70 kPa. At the arch springing monitoring point, the contact pressure decreased rapidly 6 weeks after invert pouring and then remained relatively stable by the 7th week. At day 50, due to the pouring of the second segment of the arch wall secondary lining, the contact pressure at the arch springing point showed a significant increase, finally stabilizing at 20 kPa. This process is consistently corroborated by the temporal stress curves of reinforcement in the invert concrete, which display an upward-opening parabolic shape. The development of rebar stress includes a phase sensitive to self-weight stress—not caused by surrounding rock pressure. After the surrounding rock pressure is transmitted to the secondary lining through the initial support, it similarly undergoes a process of accelerated deformation—slow deformation—stabilization. The slow deformation phase is primarily governed by the plastic and rheological properties of soft rock. Throughout the entire monitoring period, deformations of all components were mostly controlled within 20 mm, though this deformation constitutes only a minor portion. Monitoring data indicate that large deformations in soft rock result in highly asymmetric loading on the invert, with opposite stress states observed in the reinforcement. The overall reinforcement in the secondary lining is in tension, with significantly higher stresses in the crown area compared to other sections. Therefore, relying solely on structural reinforcement in the secondary lining is insufficient to resist the tensile stresses. Consequently, providing adequate tensile reinforcement to prevent cracking due to tension in the secondary lining is essential.
4. Discussion
4.1. Influence of Loose Circle in Red-Bed Soft Rock on Anchor Rod Force
The presence of a loosened zone in soft rock [33], particularly red-bed soft rock, results in pronounced asymmetry and anisotropy in the mechanical behavior of rock bolts. Monitoring has revealed that rock bolts experience higher stresses within the loosened zone of the surrounding rock, while stresses in the intact rock mass are relatively lower. Significant variations in bolt loading are observed at different locations—bolts at the crown typically endure greater stresses compared to those at areas like the hunch. Furthermore, the stress distribution in rock bolts correlates with the engineering geological characteristics and heterogeneity of the surrounding rock. In the transition zone between the loosened surrounding rock and the intact rock mass, the location of maximum stress concentration in the bolts may shift.
The peak axial force of rock bolts can serve as a critical indicator for determining the thickness of the surrounding rock loosened zone. Monitoring data reveal that the axial force of rock bolts reaches its maximum within a specific depth range, with the location of this peak value approximately corresponding to the thickness of the loosened zone. For instance, studies have identified that the axial force peaks at approximately 2.5 m, suggesting a loosened zone thickness of around 2.5 m for the investigated tunnel. This relationship provides essential guidance for support design. By analyzing the variation in bolt axial forces, the extent of rock mass disturbance can be reasonably estimated, thus enabling the appropriate determination of effective bolt length and installation parameters, as illustrated in Figure 18.
The presence of a surrounding rock loosened zone significantly influences the effectiveness of rock bolt support. Within this zone, rock bolts effectively perform their anchoring function, improving the stress state and mechanical parameters of the surrounding rock through dual mechanical and physical mechanisms, thereby enhancing the bearing capacity and stability of the rock mass. However, when the loosened zone becomes excessively thick or the surrounding rock is too weak, the support effectiveness of rock bolts may be limited. In such cases, combining other support measures becomes necessary to ensure the structural integrity and safety of the tunnel.
4.2. Influence of Support Timing in Red-Bed Soft Rock on Primary Support
The timing of support installation significantly influences the development of surrounding rock deformation and the plastic zone [34]. As shown in Figure 19, monitoring reveals that appropriately timed support can promptly control rock deformation, reduce displacement magnitude, and limit the expansion of the plastic zone radius, thereby effectively preserving rock integrity and enhancing tunnel stability. Conversely, delayed support installation may lead to excessive rock deformation, potentially resulting in softening or residual zones. This increases the structural load on the support system and elevates construction risks.
The stress state of support structures is also closely related to the timing of support installation. Appropriate timing enables the support structure to become effective during the initial stages of surrounding rock deformation, allowing it to assume partial loading promptly and preventing structural failure caused by excessive surrounding rock pressure. Conversely, delayed support installation forces the structure to withstand greater rock deformation and pressure, potentially leading to stress concentration and damage within the support system. This compromises its support effectiveness and reduces its service life.
The rational timing of support installation enables full utilization of the surrounding rock’s self-bearing capacity, facilitating cooperative interaction between the support structure and the rock mass to establish a stable support system. Research indicates that rigid support structures such as steel arch supports, when installed promptly at appropriate stages, can more effectively control surrounding rock deformation while enhancing support performance and safety. Conversely, improper timing of support installation may lead to unreasonable stress distribution within the support structure and excessive deformation of the surrounding rock, thereby increasing safety hazards during construction.
4.3. Influence of Lithology of Red-Bed Soft Rock on Support Structures
The causes of the sudden increase in initial stress
As shown in Figure 20, for water-induced softening and strength deterioration, red-bed argillaceous rock, rich in clay minerals such as montmorillonite and illite, experiences disruption of its cementation structure upon water contact. This leads to significant reduction in rock strength and a sharp decline in the self-bearing capacity of the surrounding rock [35]. After excavation, when exposed to air or construction water, the surrounding rock rapidly absorbs moisture and softens. This causes accelerated deformation within a short period, subjecting initial support elements such as rock bolts and steel arches to abrupt load increases.
Rapid release of swelling pressure: Initial rapid moisture infiltration causes immediate volumetric expansion in the surrounding rock when clay minerals absorb water, generating instantaneous swelling pressure. This process, compounded by stress redistribution effects, further drives a sharp increase in support pressure.
2.. The mechanism of subsequent pressure gradual change
Rheological behavior and long-term swelling effects: Red-bed argillaceous rock exhibits significant rheological properties, with deformation developing gradually over time under sustained loading [36,37]. Simultaneously, the swelling process of clay minerals is not instantaneous; rather, pressure is released gradually as moisture slowly diffuses, causing the rate of increase in support pressure to slow down, as shown in Figure 21.
Coordinated adjustment between surrounding rock and support: After initial support structures such as shotcrete and steel arches constrain rock deformation, the surrounding rock gradually establishes a new stress equilibrium. This slows the expansion of the plastic zone, thereby reducing the increment in support pressure.
3.. The correlation between stratum characteristics and monitoring phenomena
Low softening coefficient: The saturated strength of the rock is significantly lower than its dry strength (softening coefficient < 0.75). Rapid softening upon water exposure after excavation forces initial support structures to bear increased loads during the sharp strength reduction phase, manifested as short-term peaks in rock bolt stress and steel arch compression.
Significant expansibility: Water absorption and swelling of clay minerals not only directly increase the volume of surrounding rock but also exert pressure on support structures. The initial swelling pressure releases rapidly (e.g., abrupt stress increase in left hunch bolts), while long-term expansion—limited by the moisture diffusion rate—results in gradual pressure stabilization (evidenced by reduced fluctuations in later monitoring curves).
Anisotropy and heterogeneity: Variations in cementation degree and mineral distribution within red-bed argillaceous rock lead to spatially nonuniform softening and swelling (e.g., differential loading between right arch shoulder and left hunch). This is reflected in the “asymmetric” distribution of support pressure and localized anomalies, as illustrated in Figure 5.
4.4. The Limitations of This Study
Although the research results of this paper provide certain guidance for the construction of red-bed soft rock tunnels, there still exist limitations such as sensor monitoring errors; human errors; environmental interference; and insufficient quantitative analysis for factors like high ground stress and strain, temperature, and groundwater seepage. Future research needs to combine multi-field coupling simulation with full-section monitoring to deepen the mechanistic study.
5. Conclusions
This study, relying on the Huizhou Tunnel No. 1 and No. 2 projects, employed a wireless remote monitoring system to conduct real-time and systematic monitoring of the deformation and stress of support structures (rock bolts, primary lining, and secondary lining) in red-bed soft rock tunnels. The aim was to reveal the mechanical behavior characteristics and action mechanisms of the support system in newly constructed red-bed soft rock tunnels, providing data support for tunnel engineering under complex geological conditions. The main findings of this study are as follows: Regarding the stress characteristics of rock bolts, the stress response was sensitive, with peak values appearing rapidly and a long stabilization period (exceeding 90 days). The axial stress ranged from 105 to 330.5 MPa (with an average of 233.2 MPa at the inner end), and the force distribution showed significant asymmetry and anisotropy. Stress concentration was prominent in the crown and haunch areas, corresponding to a surrounding rock loosened zone thickness of 2.5 m, reflecting the time-dependent effect of soft rock deformation. For the primary lining mechanism, steel arches and shotcrete exhibited significant synergistic effects. The former rapidly constrained early-stage deformation (compressive stress of steel arches ≤305.3 MPa), while the latter enhanced rock stability through triaxial stress. The support timing needed to balance the self-bearing capacity of the surrounding rock and support loads; steel arches outperformed lattice girders in displacement control but bore higher loads. As for the secondary lining characteristics, contact pressure peaked during formwork removal, with the measured safety margin exceeding that specified by codes. The pressure adjustment at the invert was complex, and localized stress concentration induced significant tensile stresses in the reinforcement, necessitating adequate tensile reinforcement. The rheological properties of the surrounding rock and geological anisotropy caused spatiotemporal heterogeneity in support stress, while the weakly cemented and water-sensitive nature of soft rock intensified the structure–surrounding rock coupling, requiring dynamic monitoring to optimize support parameters.
This study confirmed the collaborative load-bearing mechanism between rock bolts and surrounding rock, systematically expounding the stress laws and action principles of support structures in red-bed soft rock tunnels. Based on the measured data from the Huizhou Tunnels, it provides a scientific basis for the design and construction of tunnels in complex red-bed soft rock areas, particularly offering engineering guidance for controlling large surrounding rock deformations (e.g., targeted reinforcement of the 2.5 m loosened zone) and optimizing the collaborative model of composite linings (load distribution of steel arches and reinforcement configuration of secondary linings), thus enhancing the safety and economy of soft rock tunnel construction.
Validation, J.W.; Investigation, J.W.; Data curation, J.W., Z.H., and F.P.; Writing—original draft, J.W. and Z.H.; Writing—review and editing, J.W., Y.W., G.C., and J.J. All authors have read and agreed to the published version of the manuscript.
The data are contained within this article.
The authors express their gratitude to the Xin Jiang Key Lab of Building Structure and Earthquake Resistance, Xinjiang University, and the Tianchi Talent Plan of the Xinjiang Uygur Autonomous Region.
The authors declare no conflict of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Figure 1 Distribution of red beds.
Figure 2 Typical Neogene red-bed soft rocks; (a) Huizhou No. 1 tunnel site area; (b) Huizhou No. 2 tunnel site area.
Figure 3 Cross-sectional diagram of measuring points and instrument layout.
Figure 4 Time history curves of stress and strain for primary lining anchor rods at different positions in the monitoring section of Huizhou No. 1 tunnel. (a) The anchor rods for the primary lining vault. (b) The anchor rods for the right arch waist of the primary lining. (c) The anchor rods for the left arch waist of the primary lining. (d) The anchor rods for the right arch shoulder of the primary lining. (e) The anchor rods inside the left arch shoulder of the primary lining. (f) The middle part of the anchor rods in the left arch shoulder of the primary lining.
Figure 5 Distribution map of anchor rod stress along the tunnel perimeter in Huizhou No. 1 tunnel (10−2 MPa).
Figure 6 Stress variation diagram of primary lining rebar gage anchor rods at the outer, inner, and middle positions. (a) Stress at the outer end of anchor rods in the section primary lining. (b) Strain in the middle of anchor rods in the section primary lining. (c) Stress at the inner end of anchor rods in the section primary lining.
Figure 7 Stress monitoring of primary lining steel arch frames in Huizhou No. 2 tunnel.
Figure 8 Surface strain of primary lining steel arch frames in Huizhou No. 2 tunnel.
Figure 9 Stress of initial shotcrete in Huizhou No. 2 tunnel.
Figure 10 Reinforcement stress monitoring of primary lining in Huizhou No. 2 tunnel.
Figure 11 Temperature variation at measurement points of rebar gages at each embedded location: (a) vault; (b) left arch waist; (c) right arch waist; (d) left arch shoulder; (e) right arch shoulder.
Figure 12 Temperature–stress relationship within left arch haunch rock bolts.
Figure 13 Contact pressure between primary lining and secondary lining.
Figure 14 Time history curves of concrete strain in secondary lining.
Figure 15 Time history curves of reinforcement stress in secondary lining concrete.
Figure 16 Reinforcement stress in invert concrete.
Figure 17 Strain of invert concrete and contact pressure with primary lining: (a) surface strain; (b) contact stress with the primary lining.
Figure 18 Influence of loose zone in red-bed soft rock on anchor rod force.
Figure 19 Influence of support timing in red-bed soft rock on primary support: (a) timely support timing; (b) untimely support timing.
Figure 20 Rock mass softening and swelling due to water infiltration.
Figure 21 Stress redistribution occurs simultaneously with the stabilization of the surrounding rock.
Statistical table of borehole acoustic logging results.
| Borehole Number | Lithology | P-Wave Velocity of Rock Mass (m/s) | P-Wave Velocity of Rock Block (m/s) | Integrity Coefficient Kv | Integrity | |
|---|---|---|---|---|---|---|
| Interval Value | Average Value | |||||
| ZK-SD-29 | Argillaceous sandstone | 1761~1923 | 1855 | 4047 | 0.21 | Highly fragmented |
| Glutenite | 2439~2924 | 2644 | 4047 | 0.43 | Moderately fragmented | |
| Conglomerate and glutenite | 2577~4274 | 3642 | 4047 | 0.70 | Moderately intact | |
| Strongly weathered glutenite | 1891~2007 | 1944 | 3148 | 0.38 | Moderately fragmented | |
| Moderately weathered glutenite | 1908~2119 | 2027 | 3148 | 0.41 | Moderately fragmented | |
| Slightly weathered glutenite | 2762~4273 | 3800 | 4200 | 0.82 | Intact | |
| Slightly weathered argillaceous sandstone | 3067~4166 | 3494 | 3938 | 0.79 | Intact | |
| SZK-SD-08 | Strongly weathered glutenite | 2604~3521 | 3001 | 1091 | 0.54 | Moderately fragmented |
| Slightly weathered glutenite | 2631~3521 | 3284 | 4091 | 0.61 | Moderately intact | |
Comparison between online health monitoring and traditional manual inspection.
| Comparison Parameters | Traditional Manual Monitoring | Online Health Monitoring |
|---|---|---|
| Timeliness | It is difficult to ensure data stability, especially in harsh weather conditions. | Real-time monitoring unaffected by weather, ensuring data stability even in harsh environments. |
| Continuity | Conduct regular inspections. | Conduct long-term uninterrupted 24 h online testing to reflect subtle change trends. |
| Accuracy | Both systematic errors and random errors are relatively large. | Basically, overcome errors caused by human subjectivity. |
| Quantifiable | Mainly based on observation, with difficulties in data quantification. | Monitor with scientific data, based on quantification, and provide massive data. |
| Convenience | Extremely tedious, requiring manual recording and then inputting into the computer. | Real-time access, backend operation, achieving automation, remote management, retrievability, and strong replicability. |
| Safety | Requires manual inspection, making it difficult to ensure human safety in harsh environments. | Safe and stable, with minimal subjective errors. |
| Economy | Each monitoring trip requires traveling to the project site, resulting in high costs for labor, travel, and significant investment for each monitoring session. | No need to travel to the project site after installation, with one-time investment for long-term use and strong economy. |
Tunnel surrounding rock monitoring items and number of measuring points.
| Monitoring Items | Equipment Name | Specifications | Equipment Model | Monitoring Location | Quantity | Equipment Images |
|---|---|---|---|---|---|---|
| Surrounding rock pressure | Pressure cell | Dimension: ∅163 × 25 mm; | FS-TY08/20 (JMZX-5020Am) | 1.2 MPa arch crown | 1 | [Image omitted. Please see PDF.] |
| 0.8 MPa arch shoulder | 2 | |||||
| 0.8 MPa arch waist | 2 | |||||
| 1.2 MPa arch springing | 2 | |||||
| 1.2 MPa middle of inverted arch | 1 | |||||
| Pressure between support and secondary lining | 1.2 MPa arch crown | 1 | ||||
| 0.8 MPa arch shoulder | 2 | |||||
| 0.8 MPa arch waist | 2 | |||||
| Internal and external forces of steel frame | Surface strain gauge | Dimension: 174 × 26 × 38 mm; | FS-BM30 | Arch crown | 1 | [Image omitted. Please see PDF.] |
| Arch shoulder | 2 | |||||
| Arch waist | 2 | |||||
| Arch springing | 2 | |||||
| Middle of inverted arch | 1 | |||||
| Secondary lining stress | Rebar meter | Measurement range: 200 MPa; | FS-B-LG10 | Arch crown | 1 | [Image omitted. Please see PDF.] |
| Arch shoulder | 2 | |||||
| Arch waist | 2 | |||||
| Arch springing | 2 | |||||
| Middle of inverted arch | 1 | |||||
| Primary shotcrete strain | Embedded strain gauge | Dimension: 174 × 26 × 38 mm; | FS-NM30 | Arch crown | 1 | [Image omitted. Please see PDF.] |
| Arch shoulder | 2 | |||||
| Arch waist | 2 | |||||
| Rock mass temperature and humidity | Thermohygrometer | Range: 0–100% | YTDY0102 | Arch crown | 1 | [Image omitted. Please see PDF.] |
| Arch shoulder | 1 | |||||
| Arch waist | 1 |
Comparative performance of steel sets vs. lattice girders.
| Parameter | Steel Sets | Lattice Girders | Difference |
|---|---|---|---|
| Peak Stress (MPa) | 178.6 (Crown) | 112.3 (Crown) | +59% |
| Deformation Rate (mm/day) | 0.85 (Pre-support) → 0.15 | 1.02 → 0.32 | −56% faster control |
| Stabilization Time (days) | 28 | 42 | −33% shorter |
| Load per Unit (kN/m) | 124.5 | 78.2 | +59% |
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Red-layer soft rock has characteristics such as softening when encountering water, loose structure, and significant rheological properties. In tunnel engineering, it is necessary to sort out and analyze the stress characteristics of its support structure. This paper focuses on the mechanical behavior and support effect during the construction of Neogene red-layer soft rock tunnels. Through field monitoring, it explores the mechanical characteristics of Huizhou Tunnel under complex geological conditions in depth. This study adopted a remote wireless monitoring system to conduct real-time monitoring of key indicators including tunnel surrounding rock pressure, support structure stress, and deformation, obtaining a large amount of detailed data. An analysis revealed that the stress experienced by rock bolts is complex and varies widely, with stress values between 105 and 330.5 MPa. The peak axial force at a depth of 2.5 m reflects that the thickness of the loosened zone in the surrounding rock is approximately 2.5 m. The compressive stress in the steel arches of the primary support does not exceed 305.3 MPa. Shotcrete effectively controls the surrounding rock deformation, but the timing of support installation needs careful selection. The stress in the secondary lining is closely related to the primary support. The research findings provide an important theoretical basis and practical guidance for optimizing the support design of red-bed soft rock tunnels and enhancing construction safety and reliability.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
; Han Zhize 1 ; Wang Yunxing 2 ; Peng, Feng 1 ; Cheng, Geng 1 ; Jia Jiaxin 1 1 School of Civil Engineering and Architecture, Xinjiang University, Urumqi 830047, China; [email protected] (Z.H.); [email protected] (F.P.); [email protected] (G.C.); [email protected] (J.J.)
2 Institute of Geological Hazards Prevention, Gansu Academy of Sciences, Lanzhou 730030, China; [email protected]