Abstract
Long-term assessment of aging highway tunnels requires linking spatial geometry with time-dependent rock-lining interactions. This study examined the 1147 m left tube of the Jinan Tunnel after 23 years of service to identify deformation-sensitive zones and inform maintenance. The tunnel is excavated mainly in weathered Cambrian limestone, classified as Grade IV (poor) to Grade V (very poor) rock. Terrestrial laser scanning (TLS) point clouds were compared with the design profile to quantify radial deviations and clearance convergence. A representative Grade IV section with an approximately 100 m overburden was simulated for 100 years using Fast Lagrangian Analysis of Continua in Three Dimensions (FLAC3D) and a Burgers viscoelastic model. At LK32+370, the crown deviation reached 109 mm, compared with 87.8 mm at the right arch waist and 46.3 mm at the left sidewall. Alignment-scale average radial deviations were approximately 40–50 mm in central sections but 60–70 mm near the ends, and Grade V sections exhibited average deviations that were approximately 1.5 times those of Grade IV sections. Modeled crown displacement increased from approximately 1.6 to 2.6 mm between years 10 and 20. From years 5 to 50, lining horizontal stress at the crown increased from 5.1 to 8.5 MPa and lining vertical stress at the arch waist from 5.5 to 10.5 MPa, with the steepest growth during years 10–20 and slower growth thereafter. The resulting framework integrates full-profile TLS screening with creep-based interpretation to prioritize poor-rock, shallow-cover, weak-interlayer, and vulnerable lining zones. Because the TLS survey represents one late-life epoch, the field quantities are design-referenced deviations rather than pure 23-year deformation.
1. Introduction
Highway tunnels are indispensable components of modern transport networks, and the management challenge is increasingly shifting from construction-led expansion toward life-cycle performance and resilience. By the end of 2025, China had 30,502 highway tunnels with a cumulative length of 35.7659 million m [1], while an increasing share of the network had entered long-term service [2]. Prolonged operation exposes tunnel structures to the cumulative effects of geological variability, groundwater, environmental action, material deterioration, and repeated traffic loading. A survey of 90 in-service highway tunnels reported lining cracks and water leakage in 81% and 69% of the investigated cases, respectively [3]. Voids, insufficient lining thickness, material degradation, spalling, and geometric distortion may coexist and interact, creating progressive pathways from localized defects to structural performance deterioration [4,5,6]. In response, tunnel condition assessment has evolved from predominantly manual inspection toward nondestructive testing, structural health monitoring, and multi-index evaluation, with combined weighting, fuzzy assessment, and data-driven methods increasingly used to improve the objectivity of condition ratings [4,5,6,7]. Nevertheless, most existing approaches still characterize visible or detectable defects and assign static condition grades. Such assessments provide limited insight into whether deformation remains active, how loads continue to redistribute between the rock mass and the lining, or whether the coupled rock mass–lining system has approached a stable state.
These questions are particularly important for tunnels excavated in weak, fractured, or intensely weathered rock. Following excavation, jointed rock, weak interlayers, and low-strength materials may undergo sustained viscoelastic or viscoplastic deformation, allowing the plastic zone, ground pressure, and lining response to continue evolving after construction [8,9,10,11]. Rock-mass heterogeneity can generate spatial differences in the deformation rate and direction and may produce asymmetric convergence [9], whereas groundwater and drainage conditions can modify effective stress and the mechanical state of the rock mass–lining interface [11]. Delayed tunnel deformation may arise from mechanical creep, seepage-induced consolidation, or a combination of both. Although these processes can produce similar time-dependent displacement signatures, they are governed by different parameters and require different mitigation strategies [11,12]. Model tests of weak rock–lining systems under operational conditions have further shown that overburden depth and the lateral pressure coefficient affect the evolution of contact pressure, lining internal forces, and displacement [13]. Rock-mass quality, weathering, weak interlayers, groundwater, and overburden should therefore be treated as interacting controls on long-term load transfer and the spatially variable tunnel response rather than as independent condition indices.
Rheological constitutive models provide a mechanical framework for representing these time effects. The classical Burgers model combines Maxwell and Kelvin elements and uses physically interpretable elastic and viscosity parameters to reproduce instantaneous elasticity, primary creep, and steady-state creep. Its relatively simple formulation and compatibility with finite-difference analysis have led to its widespread application to the long-term response of tunnels in weak rock [10,14]. Displacement-based back-analysis methods have subsequently been used to infer creep parameters and initial stress conditions from field monitoring data [10,11], while nonlinear viscous components, multiscale damage formulations, and anisotropic descriptions have extended the representation of nonlinear soft-rock deformation [14,15]. Because the standard Burgers model cannot explicitly reproduce damage-dominated tertiary creep, isotropic damage variables, fractional-order components, and viscoplastic elements have been introduced to simulate damage accumulation and accelerating deformation [16,17]. Particle swarm optimization and machine learning methods have also been explored for long-term deformation prediction and parameter identification [18]. These developments improve model flexibility but introduce additional parameters that may be difficult to identify at an engineering scale. Predictions remain sensitive to the initial stress field, groundwater conditions, rock-mass heterogeneity, rock–lining contact assumptions, and the provenance of constitutive parameters [12,15,16,17]. In the absence of long-term stress monitoring and laboratory- or field-based parameter inversion, numerical models are therefore more defensible as tools for interpreting dominant deformation modes, staged evolution, and load-transfer trends than as exact predictors of absolute stresses and displacements throughout an entire tunnel.
In parallel with advances in mechanical modeling, tunnel monitoring has progressed from discrete measurements toward full-profile, digital, and increasingly automated inspection. Leveling, convergence meters, and total stations provide accurate time histories at selected points but can miss localized or asymmetric deformation between instruments. Terrestrial laser scanning (TLS) rapidly acquires dense, non-contact point clouds and enables continuous characterization of radial profile deviations, clearance convergence, and longitudinal variations [19,20,21]. Relatively mature workflows have been developed for multi-station registration, tunnel-axis reconstruction, cross-section extraction, profile fitting, and automated deformation calculations [21,22,23]. Mobile laser scanning (MLS) and self-propelled platforms further improve the efficiency of long-distance surveys and have been applied to convergence measurement, segment-joint dislocation, and deformation analysis in shield tunnels and underground roadways [22,24,25]. Field studies in tunnels constructed using the New Austrian Tunneling Method have demonstrated the feasibility of capturing full-space deformation using three-dimensional laser scanning [26], while robot-mounted light detection and ranging and multitemporal point-cloud registration have increased the potential for repeatable long-term inspection [27]. More recent research has integrated soft-rock deformation analyses, point-cloud semantic segmentation, spatial association of defects, and numerical simulations [28,29,30]. However, when a single late-life survey is referenced only to the design profile, the resulting quantity is a present-day profile deviation that may combine construction tolerances, initial geometric differences, and service-stage changes. Point clouds alone cannot directly determine lining stress, reconstruct a historical deformation trajectory, or establish rock creep as the cause of the measured geometry.
Overall, substantial progress has been made in tunnel defect assessments, rheological analyses, and full-profile monitoring, but these evidence streams remain incompletely connected. Defect surveys primarily describe structural conditions, laser scanning resolves spatial geometry, and creep modeling addresses temporal and mechanical processes. Existing integrated studies have largely focused on construction-stage soft-rock tunnels, mine roadways, grouted ground, or geometrically regular shield tunnels, frequently using localized or relatively short monitoring records [10,14,18,25,26,27,28,29,30]. For mountain highway tunnels that have operated for more than two decades and exhibit pronounced variations in rock-mass quality and overburden, evidence linking alignment-scale profile information to the three-dimensional long-term response of the rock mass–lining system remains limited. To address this gap, the present study investigates the Jinan Tunnel on the Jinan Ring Expressway, which has been in service for approximately 23 years. High-density full-profile data from the 1147 m left tunnel were acquired using a Leica P30 terrestrial laser scanner to quantify radial deviations and clearance convergence relative to the design profile and to examine their spatial relationships with rock-mass grade and overburden. A representative three-dimensional finite-difference model was then established in Fast Lagrangian Analysis of Continua in Three Dimensions (FLAC3D), with the rock mass represented by Burgers viscoelastic rheology, to examine the staged evolution of displacement and stress in the rock mass–lining system. The field-observed spatial patterns were compared with the simulated temporal trends to identify structural positions and adverse geological combinations requiring prioritized monitoring. The working hypothesis is that the time-dependent deformation of weak and weathered rock progressively modifies load sharing between the rock mass and the lining, while rock-mass quality and overburden jointly govern the spatial nonuniformity of this response. By linking field geometry, time-dependent mechanics, and a risk-oriented interpretation, this study aims to provide a traceable evidence base for service-performance assessment and zoned maintenance of in-service highway tunnels.
2. Engineering Background
As shown in Figure 1, the research object of this study is the Jinan Tunnel on the Jinan Ring Expressway in Jinan, Shandong Province. The tunnel is a separated twin-tube structure designed for four lanes in two directions, with a clear width of 10.82 m and a clear height of approximately 8.46 m for each individual tube, as shown in Figure 2. The tunnel was completed and opened to traffic in 2002 and has been in operation for approximately 23 years. In this study, the left-line tunnel is selected for analysis. The chainage of the left-line tunnel ranges from LK32+359 to LK33+506, with a total length of approximately 1147 m. The strata traversed by the tunnel are dominated by Cambrian medium- to thick-bedded limestone. Owing to long-term weathering and fragmentation, the surrounding rock is mainly classified as Grade IV–V, with locally interbedded weak layers such as thin argillaceous rock and limestone bands. The burial depth varies considerably along the tunnel alignment, with relatively shallow cover at the entrance and exit sections and a maximum burial depth of up to approximately 100 m in the middle section.
Figure 1.
Tunnel location and geological setting: (a) satellite view and tunnel alignment; (b) location of Jinan in China; (c) geological conditions; and (d) tunnel portal. Note: In panel (c), IV and V denote surrounding-rock Grades IV and V, respectively.
Figure 2.
Schematic diagram of designed cross-section.
As shown in Figure 3, the tunnel exhibits several defects after more than two decades of service, including lining cracks, leakage, and localized behind-lining anomalies. Figure 3a shows representative circumferential/longitudinal cracking; in the annotation, δmax denotes the maximum visible crack width (approximately 5 mm), and L denotes the visible crack length (approximately 20 m). Figure 3b demonstrates that seepage traces are clearly observable at the tunnel crown and sidewalls. The behind-lining anomaly shown in Figure 3c was identified by ground-penetrating radar (GPR). GPR transmits high-frequency electromagnetic waves and interprets reflections generated by dielectric contrasts; in road and tunnel engineering, it is used non-invasively to investigate layer thickness, interfaces, moisture-related contrasts, debonding, and possible voids [31]. In the present tunnel, strong, discontinuous reflection patterns at the arch shoulder and arch waist were interpreted as GPR-indicated cavities or poorly compacted zones behind the lining. The presence of these defects suggests that the tunnel structure may have undergone long-term deformation and stress redistribution, which should be evaluated through monitoring and analysis. To further investigate the deformation characteristics of the tunnel, comprehensive field inspection and deformation measurements were conducted on the existing tunnel. During the measurement campaign, a three-dimensional laser scanner was employed to obtain high-precision cross-sectional deformation data, which will be used to evaluate the long-term deformation characteristics of the tunnel and provide a basis for the calibration of numerical simulations.
Figure 3.
Diseases of existing tunnel. (a) Lining cracks, (b) Leakage water, (c) Lining cavity.
3. Materials and Methods
3.1. Overall Workflow
Figure 4 outlines the technical route adopted in this study. First, the current operating condition of the Jinan Tunnel was investigated to establish the engineering background, including tunnel geometry, geological conditions, overburden, service history, and representative structural defects. Second, on-site full-section three-dimensional laser scanning was performed using the Leica HDS system, and the resulting point clouds were registered and processed to quantify design-referenced cross-sectional deviations and clearance convergence. Third, a representative three-dimensional FLAC3D model was developed to simulate the stress–deformation response and long-term rheological evolution of the surrounding rock–lining system, thereby evaluating the mechanical plausibility of the deformation-sensitive locations and patterns identified by the field survey. Finally, the spatial characteristics obtained from field inspection were integrated with the temporal trends and mechanical interpretations derived from the numerical simulation. The two evidence streams were cross-validated at the levels of deformation pattern, sensitive structural position, and governing mechanism rather than through a direct point-by-point comparison of absolute measured and simulated values.
Figure 4.
Technical route.
3.2. TLS Acquisition and Processing
A Leica P30 terrestrial laser scanner was used to acquire continuous full-profile point clouds along the tunnel axis (Figure 5a). Survey stations were spaced at approximately 40 m, and 360° scans covered the adjoining tunnel surfaces at each station (Figure 5b). Cyclone software was used for multi-station registration and stitching to produce the complete point-cloud model (Figure 6). The registered tunnel coordinate system and the design cross-section were aligned by chainage and section geometry. At each evaluated cross-section, the radial distance between the scanned lining surface and the design contour was calculated; the absolute values are reported when deformation magnitude is discussed. Horizontal survey lines at several elevations were also used to calculate clearance convergence as the difference between the design width and the scanned left-right clearance. The outputs were characteristic-position deviation, maximum deviation, section-averaged deviation, affected area, and clearance convergence. This is a single-epoch, late-life TLS survey referenced to the design profile. Accordingly, the calculated difference can include construction tolerance, initial as-built geometry, repair history, and service-stage change; it is not a direct time-series measurement of 23-year deformation.
Figure 5.
Three-dimensional laser scanner equipment and operating status. (a) Leica P30 3D laser scanner, (b) Full-section scan.
Figure 6.
Three-dimensional point-cloud stitching of existing tunnels.
3.3. Numerical Model and Constitutive Framework
A three-dimensional numerical model was established using the FLAC3D finite-difference software in this study. A typical section with a burial depth of H = 100 m and Grade IV surrounding rock was selected as the representative case for analysis. The tunnel axis was defined as the Y-axis. The computational domain extended 35 m on each side of the tunnel axis, forming a transverse surrounding rock domain of 70 m × 70 m. The calculated tunnel length was set to 4 m. The lining was simulated using structural shell elements, with an equivalent thickness of 0.60 m. The boundary conditions were defined as follows: horizontal constraints were applied on the lateral boundaries, a vertical constraint was applied at the bottom boundary, and the top boundary was treated as a free surface.
The Mohr–Coulomb elastoplastic model was used to establish construction-stage equilibrium, whereas the Burgers viscoelastic model described time-dependent behavior during operation; the lining remained linear elastic. In the Mohr–Coulomb model, shear failure is governed by Equation (1), where τf is shear strength, σn is effective normal stress, c is cohesion, and φ is the internal friction angle.
τf = c + σntanφ
Under constant shear stress τ, the one-dimensional deviatoric creep form of the Burgers model is governed by Equation (2), where t denotes the elapsed creep time, σ0 denotes the constant uniaxial stress applied during the creep process, G1 and η1 characterize the Maxwell branch, and G2 and η2 characterize the Kelvin branch.
The elastic and Mohr-Coulomb inputs in Table 1 were adopted from the available project design and geotechnical records. The Burgers parameters used in this study are listed in Table 2. No site-specific creep tests, repeated TLS displacement histories, or long-term stress records were available for inverse identification; consequently, the Burgers parameters are adopted as baseline model inputs rather than calibrated site parameters. The numerical analysis is therefore used to interpret deformation modes, staged evolution, and load-transfer tendencies, not as a quantitatively validated prediction of the tunnel’s historical displacement or stress trajectory.
Table 1.
Elastic–plastic and structural parameters.
Table 2.
Burgers creep parameters.
4. Results
4.1. Section Convergence Deformation Characteristics
The deformation values at characteristic positions, maximum deformation, average deformation, and clearance convergence of each cross-section were adopted as evaluation indicators and were comprehensively analyzed in combination with the corresponding surrounding rock grade and burial depth parameters. The point-cloud unfolded views of typical cross-sections and representative tunnel sections are shown in Figure 7 and Figure 8.
Figure 7.
Deformation and convergence detection results of section LK32+370. (a) Tunnel deformation, (b) Convergence deformation.
Figure 8.
Tunnel deformation nephogram of sections LK32+360 ~ LK32+410.
The TLS survey shows a systematic present-day deviation pattern relative to the design geometry. The crown generally has the largest inward radial deviation, whereas the sidewalls show the smallest values; right-side deviations are commonly greater than corresponding left-side values. At LK32+370 (Figure 7), the crown deviation is 109 mm and the right arch waist is 87.8 mm, while the left sidewall is 46.3 mm. The right arch shoulder and right sidewall reach approximately 76.1 and 64.0 mm, respectively, compared with 60.6 and 46.3 mm on the left. The recurrent right–left asymmetry is consistent with an asymmetric structural/geological loading environment, but a single design-referenced survey cannot by itself distinguish service-stage creep from construction tolerance, as-built geometry, or other historical effects.
Figure 9 and Figure 10 show pronounced longitudinal variations. Near LK32+370, the section-average radial deviation is approximately 60 mm and the maximum clearance reduction is about 160 mm. The average deviation decreases toward LK32+610, where it is approximately 37 mm, and the clearance reduction is about 80 mm. From LK32+500 to LK33+000, average deviations are mostly 40–50 mm and clearance reductions are approximately 90–110 mm. Beyond LK33+100, the average deviations increase to approximately 60–70 mm and clearance reductions to approximately 150–180 mm; at LK33+450, the average deviation is approximately 70.5 mm, and the clearance reduction is about 170 mm. Thus, the largest design-referenced geometric differences occur mainly near both tunnel ends, with lower and more stable values in the deeper central portion.
Figure 9.
Axial deformation of the operating tunnel at different mileages.
Figure 10.
Axial horizontal convergence of the operating tunnel at different mileages.
Overall, the deformation distribution along the left-line tunnel is characterized by larger deformation at both ends and smaller deformation in the middle. The initial and rear sections exhibit greater deformation due to relatively unfavorable geological and burial depth conditions. In contrast, the middle section has relatively better surrounding rock conditions and greater burial depth, resulting in smaller and more stable deformation. It should be noted that the longitudinal deformation of the tunnel does not decrease in a strictly linear manner, as local geological variations can also affect deformation development. For example, near LK32+570 of the left-line tunnel, although the burial depth reaches 60 m, weak interlayers are present in the surrounding rock, and the measured convergence at the arch shoulder is slightly greater than the average value of the adjacent sections. In general, however, better surrounding rock grades and greater burial depths correspond to smaller deformation magnitudes. The deformation in the Grade V surrounding rock sections is generally greater than that in the Grade IV sections, and the deformation in shallow-buried sections is greater than that in deeply buried sections. This pattern is consistent with the increase in support load induced by long-term surrounding rock creep and loosening. Specifically, shallow weak surrounding rock, owing to its poor self-stability, gradually transfers long-term loads to the lining, thereby causing relatively large convergence deformation.
4.2. Numerical Stress Evolution
To investigate the deformation and stress characteristics of the existing tunnel during operation, a 100-year creep calculation was conducted for the tunnel in the operation period based on the above model parameters. Figure 11 and Figure 12 present the stress contour plots of the surrounding rock and lining after 10~50 years of tunnel operation. Based on the calculation results from 10 to 50 years, a significant time-dependent redistribution can be observed in the lining–surrounding rock system. The vertical stress, σzz, continues to unload at the tunnel crown, while the compressive zone at the crown gradually expands, indicating a continuous strengthening of the arching effect. A stable vertical compression path is formed beneath the arch waist, arch foot, and invert, which bears the primary vertical load. The horizontal stress, σxx, at the sidewall and arch waist decreases with time and migrates outward. The stress contours evolve from near-field peaks to smoother and more symmetric distributions, indicating that the surrounding rock–lining system gradually changes from an initially concentrated compressive state to a broader and milder compression state. Meanwhile, the difference between σzz and σxx decreases correspondingly.
Figure 11.
Stress cloud map of tunnel operation for 10 to 50 years. (a) Horizontal stress cloud map (10 years), (b) Vertical stress cloud map (10 years), (c) Horizontal stress cloud map (20 years), (d) Vertical stress cloud map (20 years), (e) Horizontal stress cloud map (30 years), (f) Vertical stress cloud map (30 years), (g) Horizontal stress cloud map (50 years), (h) Vertical stress cloud map (50 years).
Figure 12.
Stress cloud map of tunnel lining after 10 to 50 years of operation. (a) Horizontal stress of the lining structure (10 years), (b) Vertical stress of the lining structure (10 years), (c) Horizontal stress of the lining structure (20 years), (d) Vertical stress of the lining structure (20 years), (e) Horizontal stress of the lining structure (30 years), (f) Vertical stress of the lining structure (30 years), (g) Horizontal stress of the lining structure (50 years), (h) Vertical stress of the lining structure (50 years).
The calculation results indicate that the maximum vertical stress occurs at the arch waist, the maximum horizontal stress occurs at the tunnel crown and invert, and the maximum shear stress occurs at the arch shoulder. To further reveal the time-dependent stress evolution at key locations and to match the locations of field inspection points, the temporal evolution of the horizontal stress at the tunnel crown, the vertical stress at the arch waist, and the shear stress at the arch shoulder in both the surrounding rock and the lining were selected for analysis, as shown in Figure 13. At the tunnel crown, the horizontal stress in the lining increases from approximately 5.1 MPa after 5 years of operation to approximately 8.5 MPa after 50 years of operation, while the horizontal stress in the surrounding rock increases from approximately 1.3 MPa to approximately 2.1 MPa over the same period. At the arch waist, the vertical stress in the lining increases from 5.5 MPa after 5 years of operation to 10.5 MPa after 50 years of operation, while the vertical stress in the surrounding rock increases from 2.5 MPa to 3.1 MPa.
Figure 13.
Stress evolution at characteristic positions from 5 to 50 years: (a) surrounding rock; and (b) lining.
Compared with the surrounding rock, the lining bears a higher level of additional stress during long-term operation and exhibits a more pronounced response to stress growth. In terms of time-dependent evolution, the lining stress increases significantly between 10 and 20 years after tunnel operation begins, but gradually tends to stabilize after 20 years of operation. This indicates that, during the creep process, the surrounding rock not only undergoes internal stress redistribution but also progressively transfers part of the load to the lining structure, making the lining the primary load-bearing component in long-term stress adjustment. This is consistent with the fundamental mechanism by which surrounding rock rheology causes stress to gradually transfer toward the lining. It also agrees well with the field inspection results, which show pronounced crown settlement and arch waist convergence.
Therefore, during long-term operation and maintenance, the tunnel crown, arch waist, and invert–arch foot transition zones should be regarded as key areas of concern. Particular attention should be paid to the risk of lining cracking, water leakage, and voiding induced by stress redistribution and cumulative deformation.
4.3. Numerical Displacement Evolution
To reveal the long-term deformation evolution characteristics of the operating tunnel under the rheological effect of the surrounding rock, the horizontal and vertical displacement contour plots after 10, 20, 30, 50, and 100 years of operation were extracted for comparative analysis. Figure 14 presents the displacement contour plots of the tunnel after 10–50 years of operation.
Figure 14.
Displacement cloud map after 10 to 50 years of operation. (a) X-direction displacement cloud map (10 years), (b) Z-direction displacement cloud map (10 years), (c) X-direction displacement cloud map (20 years), (d) Z-direction displacement cloud map (20 years), (e) X-direction displacement cloud map (30 years), (f) Z-direction displacement cloud map (30 years), (g) X-direction displacement cloud map (50 years), (h) Z-direction displacement cloud map (50 years).
As shown in Figure 15, the horizontal displacement increases significantly at the arch waist, followed by the arch shoulder and arch foot, while the displacement at the sidewall is relatively small. This indicates that lateral convergence is mainly controlled by deformation coordination in the arch waist–arch foot transition zone. The vertical displacement is characterized by a gradual increase in settlement above the tunnel crown over time, with the maximum settlement concentrated in the crown region. Meanwhile, a slight uplift zone can be observed near the invert, indicating a time-dependent cumulative bending response of the bottom structure. This is consistent with the stress evolution described above, which reflects crown unloading and load transfer toward the arch waist–arch foot region.
Figure 15.
Modeled displacement evolution at characteristic positions from 5 to 100 years.
As shown in Figure 15, the displacement at each key location exhibits staged growth over time. The period from 10 to 20 years represents a rapid growth stage, while the growth magnitude decreases significantly from 20 to 40 years. After 40 years, the displacement increases slowly and gradually tends to stabilize.
5. Discussion
The field observations and numerical simulations provide complementary evidence for understanding the long-term response of the operating tunnel. TLS reveals the alignment-scale spatial distribution of present-day geometric deviations, whereas the Burgers model provides a mechanical interpretation of the possible time-dependent redistribution of deformation and stress at a representative section. Larger design-referenced deviations were mainly observed in shallow, Grade V, and weak-interlayer sections, while the numerical model indicated that the crown and arch-waist regions were more sensitive to creep-induced deformation and load transfer. These findings are consistent with previous numerical studies showing that rock-mass heterogeneity can promote asymmetric and persistent deformation in weak rock [9], as well as model-test results demonstrating that overburden depth and lateral confinement significantly influence rock–lining interaction [13]. They also agree with displacement-based back-analysis studies showing that time-dependent deformation in weak rock is closely associated with progressive stress redistribution between the surrounding rock and lining [10,11]. Compared with previous TLS studies focusing primarily on cross-section extraction and geometric deformation assessment [21,22,23,26], the present study further relates alignment-scale geometric characteristics to rock-mass grade, overburden, weak interlayers, and operation-stage creep behavior. The principal contribution of this study therefore lies not in reconstructing the complete 23-year deformation history, but in integrating full-profile spatial screening with a rheological mechanism model to identify deformation-sensitive zones and support risk-oriented monitoring of aged highway tunnels.
Several limitations should be considered when interpreting these results. First, the TLS survey represents a single late-life measurement referenced to the original design profile; therefore, the observed deviations may include construction tolerances, as-built geometric differences, subsequent repairs, and service-stage deformation, which cannot be separated without a temporal monitoring baseline. Second, the Burgers parameters were adopted as engineering inputs rather than calibrated or back-analyzed from site-specific displacement histories, and their uncertainty was not quantitatively evaluated. Third, the numerical analysis was conducted for a representative Grade IV section with an overburden of approximately 100 m and therefore cannot fully represent shallow portal sections, Grade V rock, weak interlayers, or abrupt longitudinal geological variations. In addition, the surrounding rock was idealized as a homogeneous continuum with full rock–lining contact, while groundwater seepage, pore pressure, traffic loading, and thermal effects were not explicitly considered. No long-term in situ stress measurements were available for direct validation of the simulated stress evolution. Accordingly, the numerical results should be interpreted primarily in terms of deformation patterns, load-transfer mechanisms, and temporal trends rather than as quantitative predictions for the entire tunnel. Future studies should incorporate repeated TLS surveys and multi-source monitoring, determine site-specific creep parameters through laboratory testing or displacement-based inversion, and evaluate the sensitivity and uncertainty of the Burgers shear modulus, viscosity parameter, initial stress conditions, and rock–lining contact properties. Extending the model to shallow Grade V and weak-interlayer sections while considering hydro-mechanical and environmental effects would further improve its predictive reliability and applicability.
At the engineering application level, a risk-based and graded maintenance strategy is recommended for unfavorable sections, particularly shallow-buried zones, Grade V surrounding rock, and segments containing weak interlayers. These sections should receive enhanced monitoring of crown settlement, arch-waist convergence, lining cracking, leakage, and potential cavities through periodic TLS surveys and targeted inspections. Where abnormal deformation or defects persist, further structural assessment should be conducted before appropriate measures, such as void grouting, drainage improvement, or local lining reinforcement, are implemented. Relatively stable Grade IV sections may be managed through routine inspection and periodic comparative monitoring. Such differentiated monitoring and maintenance can improve the early identification of deformation-sensitive zones and support more targeted intervention during long-term tunnel operations.
6. Conclusions
This study investigated the long-term service performance of the left tube of the Jinan Tunnel, which has been in operation for approximately 23 years and is predominantly excavated in weathered Cambrian limestone classified as Grade IV–V rock. Terrestrial laser scanning (TLS) was used to characterize present-day geometric deviations relative to the design profile, and a three-dimensional FLAC3D model incorporating Burgers rheology was established for a representative Grade IV section with an approximately 100 m overburden to interpret time-dependent deformation and stress redistribution. The main conclusions are as follows:
- (1)
- The TLS results reveal pronounced spatial variability in tunnel geometry. At LK32+370, the crown deviation reached 109 mm, compared with 87.8 mm at the right arch waist and 46.3 mm at the left sidewall. Along the tunnel alignment, section-average radial deviations were generally 40–50 mm in the central portion and increased to approximately 60–70 mm near the tunnel ends. Grade V sections exhibited average deviations approximately 1.5 times those of the Grade IV sections. These results identify the crown, shallow-buried sections, Grade V rock, and weak-interlayer zones as the principal deformation-sensitive locations. Because the TLS survey represents a single late-life measurement referenced to the design profile, these values should be interpreted as present-day geometric deviations rather than the net deformation accumulated over 23 years.
- (2)
- The numerical results indicate progressive creep-induced deformation and stress redistribution in the representative Grade IV section. The modeled crown displacement increases from approximately 1.6 mm at year 10 to 2.6 mm at year 20. From years 5 to 50, the lining horizontal stress at the crown increases from approximately 5.1 to 8.5 MPa, while the vertical stress at the arch waist increases from approximately 5.5 to 10.5 MPa. The most pronounced growth occurs during approximately years 10–20, followed by a progressively decreasing rate, accompanied by load transfer from the crown toward the arch-waist–arch-foot region. The locations of stronger modeled response are consistent with the deformation-sensitive positions identified by TLS, supporting a qualitative mechanical interpretation of the observed spatial pattern.
- (3)
- The principal contribution of this study is the integration of alignment-scale TLS screening with creep-based numerical interpretation to link spatial geometric characteristics with possible long-term rock–lining interactions. However, the Burgers parameters were not calibrated against site-specific time-series measurements, and long-term in situ stress data were unavailable. Therefore, the numerical results should be regarded as mechanism- and trend-oriented rather than quantitatively predictive. In particular, the simulated temporal evolution is representative of the modeled Grade IV section and should not be directly generalized to shallow Grade V or weak-interlayer sections.
- (4)
- From an engineering perspective, monitoring and maintenance should be prioritized in shallow Grade V sections, weak-interlayer zones, and locations exhibiting large crown or arch-waist deviations, cracking, leakage, or behind-lining cavities. Periodic TLS surveys, targeted GPR inspections, crack and leakage mapping, and local convergence or strain monitoring are recommended for these higher-risk zones. Where persistent deformation or defects are confirmed, a structural assessment should precede interventions such as void grouting, drainage improvement, or local lining reinforcement. Relatively stable Grade IV sections may be managed through routine inspection and periodic comparative monitoring. These findings provide a practical basis for risk-oriented monitoring and differentiated maintenance of aging highway tunnels.
Author Contributions
Conceptualization, X.T. and W.Q.; methodology, X.W.; software, Y.D.; validation, X.W. and Y.H.; formal analysis, D.X.; investigation, T.L.; resources, W.Q.; data curation, Q.X.; writing—original draft preparation, X.W.; writing—review and editing, X.T.; visualization, Y.D.; supervision, W.Q.; project administration, W.Q.; funding acquisition, X.W. All authors have read and agreed to the published version of the manuscript.
Funding
This work was supported by the National Key Research and Development Program of China (Grant No. 2024YFF0507904), the National Natural Science Foundation of China (Grant No. 52208395), and the China University of Mining and Technology (CUMT) Open Sharing Fund for Large-scale Instruments and Equipment (Grant No. DYGX-2026-83).
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
The raw data supporting the conclusions of this article will be made available by the authors on request.
Conflicts of Interest
Author Xu Wang, Yizhe Han, Tong Li, Qiuchen Xu was employed by the company Shandong Hi-Speed Infrastructure Construction Co., Ltd. and Shandong Hi-Speed South Ring Expressway Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
References
- Lu, L.H. Data-integrated transportation, intelligence-enabled future—The 2026 China Transportation Informatization Academic and Technology Annual Conference and Symposium on AI-empowered digital and intelligent upgrading of expressways held in Xi’an. China ITS J. 2026, 7, 18–23. [Google Scholar]
- Jiang, Y.; Pan, Z.; Zhang, X. Tunnel infrastructure health management: A state-of-the-art review on defect development mechanism and robot-aided inspection system. Smart Undergr. Eng. 2025, 1, 26–39. [Google Scholar] [CrossRef] [Scilit]
- Ye, F.; Qin, N.; Liang, X.; Ouyang, A.; Qin, Z.; Su, E. Analyses of the defects in highway tunnels in China. Tunn. Undergr. Space Technol. 2021, 107, 103658. [Google Scholar] [CrossRef] [Scilit]
- Han, W.; Jiang, Y.; Wang, G.; Liu, C.; Koga, D.; Luan, H. Review of health inspection and reinforcement design for typical tunnel quality defects of voids and insufficient lining thickness. Tunn. Undergr. Space Technol. 2023, 137, 105110. [Google Scholar] [CrossRef] [Scilit]
- Zhang, G.; Liu, G.; Lu, Z.; Yan, C.; Xu, L.; Gao, Q.; Zhou, Y. Evaluation method for health state of highway tunnel structure based on adaptive comprehensive weighting. Eng. Fail. Anal. 2024, 163, 108597. [Google Scholar] [CrossRef] [Scilit]
- Wang, T.T.; Chiu, Y.C.; Li, K.J. Index for assessing spalling in tunnel lining based on displacement monitoring and crack mapping. Tunn. Undergr. Space Technol. 2024, 153, 105975. [Google Scholar] [CrossRef] [Scilit]
- Chen, W.; Tan, X.Y.; Yang, J. Review of state-of-the-art in structural health monitoring of tunnel engineering. Smart Undergr. Eng. 2025, 1, 40–50. [Google Scholar] [CrossRef] [Scilit]
- Barla, G.; Debernardi, D.; Sterpi, D. Time-dependent modeling of tunnels in squeezing conditions. Int. J. Geomech. 2012, 12, 697–710. [Google Scholar] [CrossRef] [Scilit]
- Sainoki, A.; Tabata, S.; Mitri, H.S.; Fukuda, D.; Kodama, J.-I. Time-dependent tunnel deformations in homogeneous and heterogeneous weak rock formations. Comput. Geotech. 2017, 92, 186–200. [Google Scholar] [CrossRef] [Scilit]
- Sharifzadeh, M.; Tarifard, A.; Moridi, M.A. Time-dependent behavior of tunnel lining in weak rock mass based on displacement back analysis method. Tunn. Undergr. Space Technol. 2013, 38, 348–356. [Google Scholar] [CrossRef] [Scilit]
- Tarifard, A.; Görög, P.; Török, Á. Long-term assessment of creep and water effects on tunnel lining loads in weak rocks using displacement-based direct back analysis: An example from northwest of Iran. Geomech. Geophys. Geo-Energy Geo-Resour. 2022, 8, 31. [Google Scholar] [CrossRef] [Scilit]
- Leone, T.; Nordas, A.N.; Anagnostou, G. Creep Versus Consolidation in Tunnelling Through Squeezing Ground—Part A: Basic Time Effects. Rock Mech. Rock Eng. 2024, 57, 5519–5536. [Google Scholar] [CrossRef] [Scilit]
- Wang, P.; Ma, X.; Yang, L.; Sheng, X.; Wang, X.; Lin, C. Investigation into the time-dependent characteristics of stress and deformation of weak surrounding rock and lining structure in operational tunnels: Model test. Appl. Sci. 2024, 14, 5447. [Google Scholar] [CrossRef] [Scilit]
- Yin, Z.; Zhang, X.; Li, X.; Zhang, J.; Zhang, Q. Modified Burgers model of creep behavior of grouting-reinforced body and its long-term effect on tunnel operation. Tunn. Undergr. Space Technol. 2022, 127, 104537. [Google Scholar] [CrossRef] [Scilit]
- Xia, C.; Liu, Z.; Zhou, C.; Zhang, L. A meso/macroscale theoretical model for investigating the large deformation of soft rock tunnels considering creep and anisotropic effects. Rock Mech. Rock Eng. 2023, 56, 4901–4922. [Google Scholar] [CrossRef] [Scilit]
- Kabwe, E.; Karakus, M.; Chanda, E.K. Isotropic damage constitutive model for time-dependent behaviour of tunnels in squeezing ground. Comput. Geotech. 2020, 127, 103738. [Google Scholar] [CrossRef] [Scilit]
- Tarifard, A.; Török, Á.; Görög, P. Fractional derivative-based burger creep model for soft rocks and its verification using tunnel monitoring results and experimental data. Arab. J. Sci. Eng. 2025, 50, 16453–16464. [Google Scholar] [CrossRef] [Scilit]
- Kovačević, M.S.; Bačić, M.; Gavin, K.; Stipanović, I. Assessment of long-term deformation of a tunnel in soft rock by utilizing particle swarm optimized neural network. Tunn. Undergr. Space Technol. 2021, 110, 103838. [Google Scholar] [CrossRef] [Scilit]
- Mukupa, W.; Roberts, G.W.; Hancock, C.M.; Al-Manasir, K. A review of the use of terrestrial laser scanning application for change detection and deformation monitoring of structures. Surv. Rev. 2017, 49, 99–116. [Google Scholar] [CrossRef] [Scilit]
- Teng, J.; Shi, Y.; Wang, H.; Wu, J. Review on the research and applications of TLS in ground surface and constructions deformation monitoring. Sensors 2022, 22, 9179. [Google Scholar] [CrossRef] [Scilit]
- Xie, X.; Lu, X. Development of a 3D modeling algorithm for tunnel deformation monitoring based on terrestrial laser scanning. Undergr. Space 2017, 2, 16–29. [Google Scholar] [CrossRef] [Scilit]
- Cui, H.; Ren, X.; Mao, Q.; Hu, Q.; Wang, W. Shield subway tunnel deformation detection based on mobile laser scanning. Autom. Constr. 2019, 106, 102889. [Google Scholar] [CrossRef] [Scilit]
- Jia, D.; Zhang, W.; Liu, Y. Systematic approach for tunnel deformation monitoring with terrestrial laser scanning. Remote Sens. 2021, 13, 3519. [Google Scholar] [CrossRef] [Scilit]
- Xu, L.; Gong, J.; Na, J.; Yang, Y.; Tan, Z.; Pfeifer, N.; Zheng, S. Shield tunnel convergence diameter detection based on self-driven mobile laser scanning. Remote Sens. 2022, 14, 767. [Google Scholar] [CrossRef] [Scilit]
- Wu, C.; Li, Y.; An, L.; Dong, E.; Han, L. An MLS-based high-accuracy measurement and automatic analysis method for roadway deformation. Tunn. Undergr. Space Technol. 2023, 140, 105306. [Google Scholar] [CrossRef] [Scilit]
- Hu, D.; Li, Y.; Yang, X.; Liang, X.; Zhang, K.; Liang, X. Experiment and application of NATM tunnel deformation monitoring based on 3D laser scanning. Struct. Control Health Monit. 2023, 2023, 3341788. [Google Scholar] [CrossRef] [Scilit]
- Jeong, S.; Kim, M.G.; Park, J.Y.; Oh, K.-Y. Long-term monitoring method for tunnel structure transformation using a 3D light detection and ranging equipped in a mobile robot. Struct. Health Monit. 2023, 22, 3742–3760. [Google Scholar] [CrossRef] [Scilit]
- Pu, J.; Yu, Q.; Zhao, Y.; Li, Z.; Cao, Y.; Le, Z.; Yang, Z.; Li, X. Deformation Analysis of a Roadway Tunnel in Soft Swelling Rock Mass Based on 3D Mobile Laser Scanning. Rock Mech. Rock Eng. 2024, 57, 5177–5192. [Google Scholar] [CrossRef] [Scilit]
- Li, J.; Zhang, Z.; Sun, H.; Xie, S.; Zou, J.; Ji, C.; Lu, Y.; Ren, X.; Wang, L. GL-Net: Semantic segmentation for point clouds of shield tunnel via global feature learning and local feature discriminative aggregation. ISPRS J. Photogramm. Remote Sens. 2023, 199, 335–349. [Google Scholar] [CrossRef] [Scilit]
- Wang, X.; Wang, Y.; Yang, Y.; Fang, Y.; Zhuo, B. Distribution characteristics and spatial correlation analysis of defects in in-service metro shield tunnels: A case study. Tunn. Undergr. Space Technol. 2026, 170, 107397. [Google Scholar] [CrossRef] [Scilit]
- Slabej, M.; Grinč, M.; Kováč, M.; Decký, M.; Šedivý, Š. Non-invasive diagnostic methods for investigating the quality of Žilina airport’s runway. Contrib. Geophys. Geod. 2015, 45, 237–254. [Google Scholar] [CrossRef] [Scilit]
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. |
© 2026 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.
















