Accurately characterising the spatial variability of rock-mass properties is essential for tunnel stability analysis and risk control. However, the absence of physical constraints and the inherent smoothing of traditional interpolation techniques make it challenging to transform high-frequency, one-dimensional measurement-whilst-drilling (MWD) data into three-dimensional heterogeneous numerical models. This study proposes a data- and mechanism-driven intelligent modelling framework for refined characterisation of tunnel rock masses. A multi-stage theoretical inversion scheme based on mechanical specific energy (MSE) and the Hoek–Brown criterion is first developed to derive equivalent Mohr–Coulomb parameters (E, c, φ) from systematically cleaned MWD data. Subsequently, a hybrid spatial interpolation algorithm integrating sequential Gaussian simulation (SGS) with Hoffman correction is formulated. This approach reconstructs high-fidelity heterogeneous parameter fields that preserve the geostatistical structure whilst remaining strictly conditioned on hard drilling data. Validation against field monitoring shows that the proposed intelligent model reduces the average relative deformation prediction error from 18.4

Blasting-induced rock fragmentation in sandstone is commonplace in tunneling and mining engineering. This paper presents an experimental and numerical investigation into the disintegration characteristics and energy consumption of sandstone under blast loading. Twelve lab-scale blast tests using sandstone cylinders are conducted with three stemming configurations (no-, partial-, and full-stemming) and four charge densities. Blast-produced fragment size distributions are first obtained by sieving and weighing. The shape of fragments is analyzed by image processing, and the energy consumption is determined using the Griffith fracture criterion. Additionally, LS-DYNA simulates pressure evolution, crack development, sandstone disintegration, and energy dissipation. The results quantitatively demonstrate that rock disintegration becomes significantly finer in a sequence of no-stemming, partial-stemming, and full-stemming, with the average fragment size decreasing by approximately 18.4% and 27.9% for partial- and full-stemming configurations, respectively, compared to unstemmed blasts. The aspect ratios of fragments are primarily falling within the range of 0.6-0.8, showing no statistically significant dependence on stemming configuration (p > 0.5, rs = -0.062). The proportions of explosive energy utilized for rock fragmentation under unstemmed, partially stemmed, and fully stemmed blasting are quantified as 5.15%–8.54% (mean 6.57%), 5.90%–11.90% (mean 9.36%), and 8.28%–14.00% (mean 10.40%), respectively, with the explosive energy utilization enhanced by approximately 42.47% (partial-stemming) and 58.30% (full-stemming) compared to the unstemmed case. Numerical modelling confirms that stemming enhances confinement, prolongs pressure duration, and improves the energy transfer from explosive detonation to internal and kinetic modes, thereby leading to greater efficiency in both rock fragmentation and energy conversion.

Intense impact load may cause local damage and segmental joint deformation in shield tunnels. However, the controlling effect of the secondary lining on the impact resistance enhancement still remains unclear. In this study, the mechanisms of segment concrete damage and segmental joint deformation under intense oblique impact load were revealed. Based on the load equivalence law of energy flux density (EFD), four benchmark tests with the reduced-scaled geometric ratio of 1/3.43 were conducted considering the influence of impact locations and secondary lining thicknesses. By the experimentally validated refined finite element (FE) model, parametric analysis was conducted considering various secondary lining thicknesses and impact locations. The following conclusions can be drawn: (1) The refined FE model accurately reproduced the failure modes and residual deformation observed in oblique impact tests, with maximum relative errors of 2.53 % (opening) and 4.62 % (dislocation). (2) Propagation of damage in the segment concrete was effectively mitigated by the secondary lining. (3) An evaluation method that comprehensively considered the impact resistance and economic efficiency of the segmental lining system under oblique impact loading was established. (4) The optimal secondary lining thickness, identified as 42.86 % of the segment thickness, was derived from incremental marginal benefit assessment. (5) The secondary lining can optimize the impact load transfer path, forming an efficient cooperative bearing mechanism, with segment concrete energy absorption decreases by 32.81 %–61.54 % at the optimal secondary lining thickness.

The joint interface is a critical component of the segment lining structure. However, its current design largely depends on analogy and empirical methods, which often prevent the joints from fully utilizing their load-bearing capacities, specifically, bending stiffness and bearing capacity. To address this limitation, first, a refined 3D numerical model was developed, capable of parametrically modelling the joint and automatically determining its bending stiffness. Second, a multi-objective optimization model was formulated, incorporating both bending stiffness and bearing capacity as performance criteria. This model was solved using a genetic algorithm-based approach. Third, the effectiveness of the optimization method and the influence of key parameters were systematically investigated. Finally, refined numerical models of the segment lining structure were established to evaluate the effects of joint configuration optimization on the mechanical performance of the overall structure. The results demonstrate that the proposed optimization method is highly effective. The maximum increases in bending stiffness and bearing capacity of joints with single and double waterproof structures were 16.9% and 19.8%, and 11.0% and 11.0%, respectively. The variation ranges of bending stiffness and bearing capacity for joints with single and double waterproof structures were - 9.5-30.5% and - 5-18.3%, respectively, indicating the necessity of optimizing joint interface parameters. For joints with a single waterproof structure, both bending stiffness and bearing capacity exhibit greater improvements after optimization when the axial force is higher or the weight coefficient assigned to bending stiffness is lower. Moreover, the influence of the bending stiffness weight coefficient on the optimization outcomes diminishes as the axial force increases. For joints with double waterproof structures, axial force has a relatively minor influence on the optimization results. When both the axial force and the weight coefficients of bending stiffness are either relatively low or high, the weight coefficient exerts a more significant impact on the optimization outcomes. Optimizing the interface configuration of the segmental joint can effectively control the deformation of the lining structure, with reductions in radial displacement ranging from 13.91% to 52.92%. Additionally, the bending moment and joint opening at the localized position are significantly reduced following optimization.

Subsea shield tunnels, operating under high water pressure and aggressive chloride environments, face significant challenges in long-term safety and reliability due to progressive segmental lining degradation and uncertainty in structural performance evolution. This study proposes a life-cycle reliability assessment framework for subsea shield tunnel segmental linings, capable of quantifying time-dependent deterioration and predicting service life under uncertain mechanical and environmental conditions. The framework integrates: refined finite element model incorporating parameter randomness, life-cycle deterioration limit state equations, and improved reliability analysis approach combining the Akaike Information Criterion (AIC) with the consideration of negative reliability. An actual large cross-section, deep-buried, high-water-pressure subsea shield tunnel is analyzed to capture the spatial-temporal evolution of mechanical responses and reliability indices, followed by service life prediction based on failure probability development. Key findings reveal that: (1) Negative reliability emerges in the final 20 years of the 100-year service life, signaling the need for repair and safety enhancement measures by the mid-to-late service stage; (2) The bending moment distribution type at the vault remains notably stable compared to other locations, showing lower sensitivity to variability in geotechnical and structural parameters; (3) Bending moment variability is consistently highest at the spandrel and springline, attributed to their role as transition zones between positive and negative moments, while axial force variability peaks at the invert; (4) Parametric variability influences the statistical characteristics-but not the spatial distribution patterns-of internal forces, with variability control proving more effective in enhancing early-service reliability than improving late-service safety. The proposed framework provides a quantitative tool for reliability-based prediction, offering practical support for maintenance optimization, structural reinforcement strategies, and risk management in subsea shield tunnel engineering.

Sandy pebble strata, characterized by high pebble content, high uniaxial compressive strength (UCS) of pebbles, poorly-sorted grain size distribution, and low flowability, often induce muck accumulation, cutterhead clogging, and chamber caking during earth pressure balance (EPB) shield tunneling, thereby reducing construction efficiency and compromising safety. Existing studies, predominantly relying on macroscopic indicators, often fail to elucidate the mesoscopic particle transport behaviors and the synergistic mechanism of the cutterhead-agitator system. This study develops a full-scale, three-dimensional finite difference method-discrete element method (FDM-DEM) coupled model to simulate particle transport within the cutterhead and soil chamber, to systematically reveal the transport mechanisms and guide equipment optimization. Results indicate that: (1) the soil chamber exhibits a "core-ring" dual-flow structure, characterized by a low-velocity "dead-zone column" at the center and a high-velocity annular channel at the periphery, attributed to the limited opening ratio in the central area and insufficient agitator coverage; (2) a "barrier-channel-dead-zone column" triple-control mechanism is proposed to explain the constraints imposed by energy barriers, geometric confinement, and cascade migration on particle transport; and (3) a synergistic optimization strategy is proposed, which includes increasing the opening ratio in the cutterhead's central region, installing vertical agitators at the chamber center, and relocating the inner-ring agitator to R = 1.8 m. These findings provide significant theoretical insights for understanding muck transport behaviors and optimizing EPB shield design in sandy pebble strata.

The continuous development of karst in carbonate rock formations leads to an imbalance in the seepage field and stress field, while the crystallization blockage in tunnel drainage systems further increases the burden on the lining, seriously threatening the long-term safe operation of the tunnel. To investigate the mutual feedback mechanisms between solute transport, crystallization and tunnel seepage and stress fields during the dissolution process of carbonate rockmass, the Tuzhu Tunnel in Chongqing, China, is taken as the engineering background, on-site sampling and testing, theoretical analysis and numerical simulation methods are employed, a theoretical framework and a three-dimensional finite element model coupling chemical reactions, seepage fields and stress fields are established. The evolution characteristics of surrounding rock porosity, water pressure and Ca2+ concentration under long-term CO2 erosion are revealed, and the concentration distribution of crystalline deposits in the primary support and drainage system, as well as the evolution characteristics of lining load and damage are analyzed. The results indicate that the dissolution of surrounding rock and the evolution of the seepage field exhibit a positive feedback effect. The seepage channels provide sites for chemical reactions, while chemical erosion increases the porosity of the surrounding rock, thereby leading to an increase in water pressure. The concentration of crystalline substances in the primary support is primarily governed by flow velocity. And the concentration of crystalline substances in the primary support shows a non-uniform distribution influenced by the drainage system, with the maximum concentration stabilizing at approximately 4.3 mol/m3 over time. The concentration of crystalline substances in the drainage system decreases progressively along the water flow, reaching its peak at the vault in the circumferential drainage pipes (2.68 mol/m3) and at the connection with the circumferential pipes in the longitudinal drainage pipes (1.07 mol/m3). The crystallization amount in the drainage system, water pressure and damage factor of the lining increase nonlinearly with time. Local crystallization blockage in the drainage pipes can cause uneven stress distribution in the lining, potentially resulting in localized damage, with the most significant influence on the mechanical damage of the lining structure observed in cases of partial blockage in the longitudinal drainage pipes.

Layered phyllite, characterized by its distinct and complex mechanical properties, often presents engineering challenges during tunnel construction, including rock mass collapse and asymmetric structural failure. These issues are particularly severe under unloading conditions, where the mechanical properties of the rock deteriorate further, posing significant threats to engineering safety. This study integrates laboratory experiments, numerical simulations, and field monitoring to conduct a comprehensive investigation into the unloading mechanical properties of layered phyllite, applying the findings to real-world engineering projects. The study reveals that layered phyllite exhibits significant differences between conventional triaxial loading tests and unloading tests, with the rock exhibiting a tendency toward brittle failure and larger stress drops during unloading. The mechanical properties of layered phyllite specimens are highly sensitive to changes in the unloading rate; increased unloading rate can intensify stress concentration and energy release, which in turn reduces the peak strength and aggravates failure. However, the sensitivity of layered phyllite to initial confining pressure and bedding angle remains consistent across different unloading rates. Furthermore, numerical simulations of the unloading tests based on the SUBI (Bilinear Strain-Softening/Hardening Ubiquitous-Joint) constitutive model align closely with laboratory test results, although further parameter adjustments are required for improved accuracy. The excellent agreement between the field monitoring data and the numerical predictions from the SUBI constitutive model validates its effectiveness and accuracy in simulating the tunnel excavation response in layered phyllite. Furthermore, a comparative analysis using this validated model against the UBI-J(Ubiquitous-Joint) model, which ignores unloading effects, quantifies the significant impact of excavation-induced unloading at an engineering scale.

Reliable data sources are essential for intelligent tunnel construction, yet on-site data are often insufficient to meet sample requirements. Previous numerical modeling studies have seldom considered the combined effects of different excavation methods’ spatial effects and the distribution characteristics of joints. This paper develops a method to construct a stability database for jointed rock tunnels with primary support systems using a computational framework combining the finite difference method–discrete element method (FDM–DEM). The framework constructs a 2D model using the Mohr–Coulomb criterion and a 3D model with the Hoek–Brown failure criterion, enabling the stress release process to accurately replicate the influence of joint distribution features and excavation space effects in the 2D calculations by utilizing the longitudinal deformation profile parameters of the bench sections and fine-grained ground reaction curves. The computational circle is determined by grid research and data analysis, while the performance differences of various primary support components and their correlations with surrounding jointed rock are analyzed using the control variable method. The validity of the framework is initially confirmed by case comparisons and macroscopically validated using the Mantel test and Spearman analysis on the constructed simulation database—containing tunnel construction information, joint distribution, rock mechanics parameters, and stability indices—thereby establishing a reliable foundation for machine learning and transfer learning applications.

Tunnel-internal structure interaction under strike-slip faulting is critical for the safe design of electrical equipment in GIL utility tunnels, but its system behavior remains insufficiently understood. This study develops a three-dimensional numerical model of a fault-crossing GIL utility tunnel with internal structures and investigates a series of representative cases. The structural response and damage characteristics of the GIL utility tunnel are analyzed, the tunnel-internal structure interaction mechanisms are summarized, and the effects of longitudinal deformation joints are evaluated. Results indicate highly non-uniform coupled deformation in the GIL utility tunnel, with severe cross-sectional distortion and damage concentrated near the fault zone. Internal structures significantly modify the local cross-sectional response through enhancing stiffness, with the substructure providing stronger deformation restraint than the GIL pipelines. The tunnel-internal structure interaction is governed by longitudinal shear transfer and transverse squeezing-induced uplift, and a clear left-right asymmetry is observed in GIL pipeline responses, with the deviation mode varying with pipeline height. Longitudinal deformation joints act as kinematic buffers by absorbing shear displacement at structural discontinuities, thereby reducing cross-sectional distortion, suppressing uplift, and mitigating damage propagation. The findings provide a mechanical basis for the fault-resistant design of GIL utility tunnels.

Soil conditioning is widely used to ensure tunnel efficiency in earth pressure balance (EPB) shield tunnelling. Compressibility, shear strength and tangential adhesion strength are key parameters to evaluate the effectiveness of soil conditioning. This paper studies these parameters by means of pressurized vane and plate shear tests and compression tests. The total vertical and lateral pressure, pore pressure, peak and residual torque are recorded and analysed. The results show that the difference between the peak and residual torque of vane and plate shear tests decreases with foam injection ratio (FIR), and the injection of foam increases the ratio between tangential adhesion strength and shear strength (ratio alpha). There is a transitional void ratio (etran) of 1.1 times the maximum void ratio (emax). Conditioned sand behaves like isotropic material when the void ratio is higher than etran. Effective stress starts to show when the void ratio is lower than etran, leading to an increase of the torque measured, a decrease of compressibility and a decrease of ratio alpha from 0.76 to 0.47. The test results for foam with various foaming agent concentrations (Cf) show that the Cf of foam should be higher than 3% or the foam bubbles will collapse during the mixing with sand.

To mitigate safety risks posed by climate extremes to mountain tunnels, this study had developed an integrated physics-artificial intelligence framework to predict the hydrological-mechanical cascade from rainfall forcing to hydraulic head and lining stress. A three-dimensional groundwater flow model had been built using the modular finite-difference groundwater flow model software (MODFLOW) to generate transient hydraulic head fields under rainfall and drought scenarios. A local-scale tunnel mechanical model had been established using the finite element analysis software ABAQUS to produce lining stress samples. Two cascaded surrogate models had then been trained: a long short-term memory network coupled with a multilayer perceptron had been used to forecast heads from rainfall time series, and a categorical boosting regressor (CatBoost) had been used to estimate lining stresses from the predicted heads and key geological descriptors. MODFLOW had reproduced observed heads at five monitoring points with a root mean square error of 19.87 m and a relative error of 1.59%. Scenario analyses had shown rapid, spatially heterogeneous head rises in high-permeability zones during extreme rainfall and sustained head declines under drought. The long short-term memory network coupled with a multilayer perceptron had achieved a coefficient of determination of 0.90 for head prediction. CatBoost had predicted maximum and minimum principal lining stresses with coefficients of determination of 0.95 and 0.98, respectively, and had identified high-stress segments along the tunnel alignment. Overall, the cascade surrogate strategy had enabled near-real-time rainfall-to-risk mapping with reduced computational cost, supporting operational risk assessment and drainage-oriented decision support.

Reliable multi-step forecasting of tunnel boring machine (TBM) operational performance is essential for predictive monitoring, risk-aware decision-making, and semi-autonomous mechanized tunnelling. However, TBM telemetry is characterized by multi-scale temporal dependencies, geological-transition-induced nonstationarity, machine wear, operating-strategy shifts, and horizon-dependent error accumulation. To address these challenges, this study proposes TCN-AOL, a geology-aware attention-augmented temporal convolutional framework with online-compatible sequential adaptation for real-time TBM performance prediction. TCN-AOL combines causal feature construction, multi-dilated temporal convolution, multi-head self-attention, and a lightweight geological mixture-of-experts adapter conditioned on discrete geological states. Field experiments on a Yellow River-crossing metro tunnel demonstrate that the proposed model achieves strong short-horizon accuracy, with single-step R2=0.974 in the main rolling evaluation, and maintains R2>0.90 over the first 5–7 prediction steps under rolling observation-assimilation evaluation (h0=7). Under strict autoregressive rollout (h0=0), performance degrades faster because of accumulated prediction errors. Transition analysis, ablation studies, data-efficiency tests, window-sensitivity evaluation, robustness assessment, and latency analysis further validate the effectiveness and practical applicability of the framework. These results suggest that geology-aware temporal convolution and attention-enhanced sequential modelling provide a reliable basis for multi-horizon forecasting in intelligent tunnelling.

To address the challenge of deformation control in operating tunnel structures, this study investigates a novel reinforcement method for operating shield tunnels using a basalt-fiber-reinforced polymer-wrapped concrete-filled steel tube (BFRP-CFST) composite profile. Two primary study variables were considered. At the segment level, the reinforcement condition comprised two levels: unreinforced and BFRP-CFST-reinforced, with three replicate specimens at each level (US-1 to US-3 and RS-1 to RS-3, respectively). At the full-ring level, the number of installed composite profile frames comprised five levels (n = 0, 1, 2, 3, and 4), where n = 0 represented the unreinforced reference condition. A combined experimental and numerical framework was established, including full-scale four-point bending tests on individual tunnel segments and finite element simulations of full-ring linings. Experimental results demonstrate that the ultimate bearing capacity of reinforced segments increased from 534.4 kN to 921.3 kN, corresponding to a 72.4% improvement. The load level before visible cracking increased by 84.2%. At maximum crack widths of 0.2 mm and 2.0 mm, the mid-span displacement of the reinforced segments was reduced by 30.0% and 21.4%, respectively. The test observations indicate that the prefabricated composite profiles effectively delayed crack development and improved the post-cracking stiffness of the segment. Full-ring numerical simulations further showed that installing one to four composite profile frames increased the external load corresponding to a convergence displacement of approximately 10.5 cm by 12.4%, 21.1%, 28.5%, and 37.4%, respectively. Scientifically, the results reveal a staged load-transfer process in which adhesive bonding provides distributed load transfer during the initial response, while mechanical anchors maintain residual load transfer after local interface debonding; they also establish a quantitative relationship between the number of profile frames and full-ring convergence resistance. From an applied engineering perspective, the proposed profile increased the ultimate load and crack-initiation load of the segments by 72.4% and 84.2%, respectively, while its lightweight and prefabricated configuration provides a potentially rapid rehabilitation option for operating shield tunnels.

Blasting in sandstone rock mass widely exists in tunneling and mining engineering. In this study, the rock fragmentation and energy dissipation under sandstone blasting are experimentally and numerically investigated with the variation in powder factor (0.188–0.983 kg/m3). Six tests of blast-induced rock fragmentation are conducted using sandstone cylinders with dimensions of 200 mm in diameter and 200 mm in height. The blast-induced fragment size distributions (FSDs) are quantitatively analyzed through sieving and curve fitting. Then, the fragment geometry, including fragment size and fragment shape, is further examined via image processing, and the energy dissipation in sandstone fragmentation is calculated based on Griffith fracture theory. Furthermore, the disintegration of sandstone cylinder and associated energy consumption during blasting are modeled in LS–DYNA. The physical tests show that the rock fragmentation shifts finer and uniform as the charging density increases, leading to a significant proportion reduction in large fragments and a prominent increment in small and fine fragments. The blast-produced FSD can be well-characterized by Weibull distribution, three-parameter GEV function and extended Swebrec function, with the latter demonstrating superior accuracy in capturing both fine and coarse fragments. Meanwhile, the fragment shape becomes rounder first and then changes to more elongated. During blasting, the sandstone cylinder fracturing consumes 6.69–14.00

Conventional stress-based methods for pressure arch characterization are limited to circular tunnels under hydrostatic stress conditions, whereas numerical simulation-based methods do not adequately account for mechanized tunneling effects. To address these limitations, this study integrates model tests, numerical simulations, and field tests to investigate the pressure arch effect, establish a multi-dimensional evaluation framework, and precisely delineate the pressure arch’s boundaries under different conditions. The results demonstrate that different criteria for pressure arch determination must be adopted, as reliance on a single criterion tends to result in significant errors. Error bar analysis of discrepancies among results from four criteria reveals that, after removing obvious outliers, the error values all remain below 0.14. Increased lateral pressure coefficients exacerbate the complexity of pressure arch, necessitating different determination criteria. These research findings offer direct guidance for mechanized tunnel construction: (1) The methodology enables dynamic assessment of rock-load transfer mechanisms during excavation, providing a theoretical basis for full-face and long-cycle tunneling strategies; (2) A pressure-arch-theory-driven formulation is established to quantify surrounding rock pressures in mechanized tunnels.

Abstract This paper investigates the mechanical characteristics of a multiple‐layered lining shield tunnel structure utilized in the Water Resources Allocation Project of the Pearl River Delta through theoretical analysis and model testing. The multiple‐layered lining structure consists of an exterior precast reinforced concrete segment lining, a middle layer of self‐compacting concrete (SCC), and an interior steel pipe. With emphasis laid upon lining interactions and the effect of different burial depths, with 16 m considered as the minimum burial depth and 40 m as the maximum burial depth, the structure's performance has been evaluated during the operation. During the operation, under increasing water pressure, axial pressure decreases while tension increases, with inner linings effectively bearing the load and protecting outer segments. Therefore, deformations change; outer linings undergo greater outward displacements and fewer inward displacements. With acoustic emission (AE) counts of 2286 and 2057, respectively, in the minimum depth (16 m) and the maximum depth (40 m) loading conditions, AE monitoring shows persistent micro‐cracking, but the structure is still elastic and safe under the designed pressure of 0.75 MPa. However, to optimize long‐term performance, further research is necessary to address possible cracking and expansion. The multiple‐layered lining technique is a promising substitute for conventional methods and proves to be efficient, economical, and suitable for tunnels with moderate to high internal water pressure.

To overcome the challenges of deformation control in segmental linings of super-large diameter shield tunnels constructed in soft ground, a novel segmental joint with double connections featuring DDCI connector and oblique bolts has been proposed. In this study, full-scale tests and numerical simulations were conducted to investigate the deformation processes, failure mechanisms of the proposed joint, with particular emphasis on the mechanical behavior of the DDCI connectors and bolts. For comparison, a conventional single-connection joint equipped with oblique bolts was also investigated, and the flexural mechanical performance of the two types of joints was compared in terms of peak strength, stiffness, and ductility. The results show that the double-connection joint undergoes four distinct stages of mechanical response under positive and negative moments. The joint fails under positive moment primarily due to concrete crushing at the joint surface, while failure under negative moment stems from the fracture of the anchor rebars in the vicinity of the DDCI connector. Under positive moment, the DDCI connectors mainly experience bending, while they are subjected to stretching under negative moment. The DDCI connector and bolt contribute more significantly to the load-bearing capacity of the double-connection joint under negative and positive moments, respectively. The addition of DDCI connector significantly enhances the flexural mechanical performance of the segmental joint. Compared with the single-connection joint, the double-connection joint with DDCI connector exhibits increases of 1.10–1.25, 1.11–1.67, and 0.98–1.39 times in peak strength, stiffness, and ductility, respectively, under positive moment. Under negative moment, these enhancements further increase to 1.19–1.87, 1.18–1.89, and 1.43–2.27 times, respectively.

This study presents an experimentally informed coupled creep-damage modelling framework for the time-dependent micromechanical behaviour of hardened cement paste. Using realistic microstructures, a viscoelastic formulation solved by an exponential algorithm, and a continuum damage model, the framework consistently captures microscale creep, creep recovery, and strain-rate-dependent response. The results show that microstructural discretization strongly affects predictions: coarse discretizations underestimate porosity and overestimate hydration, leading to overpredicted stiffness and strength. The model reproduces measured flexural strength, elastic modulus, and the observed brittle failure. Low-stress creep and recovery are predicted accurately, with high recovery ratios indicating predominantly linear creep. The calibrated HD-CSH/ LD-CSH creep modulus ratio is consistent with experimental insights, supporting that calcium hydroxide increases the creep modulus of CSH. Strain-rate effects are also captured: slower loading allows more creep and earlier damage in weaker phases, while faster loading drives damage into stronger phases, yielding higher apparent stiffness and strength with more pronounced damage patterns. Overall, the framework provides a physically consistent basis for studying microscale creep-damage interactions. Future work will incorporate temperature and moisture effects and extend the approach to multiscale simulations of long-term concrete behavior under realistic environments.

The responses of transportation tunnels subjected to faulting highly depend on the fault movement features. However, it is difficult for the present physical modeling to model the different movement features of creep-slip and stick-slip faults. This study applies a recently developed numerical model to investigate the movement features of creep-slip and stick-slip faults and their effects on tunnel responses. The numerical model considers both the fault internal structure and movement features of creep-slip and stick-slip faults. The typical characteristics of creep-slip and stick-slip faults are distinguished from previous field observations and experiments, including sliding type, movement features, and slip rate. The numerical results show that the friction coefficient at the slip surface, rather than the slip rate, controls the movement features. The critical value of the friction coefficient for creep-slip and stick-slip faults is about 0.6. A faster slip rate leads to a little larger shear deformation concentrated in the damage zone. The movement features have remarkable effects on tunnel responses. Compared with the tunnel in a creep-slip fault, the tunnel experiences more concentrated longitudinal displacement, larger cross-sectional deformation, and more severe damage in a stick-slip fault. A larger frictional coefficient leads to less severe damage and longer damaged length of the tunnel in a stick-slip fault. Slip rate affects the damage process and the damaged length of the tunnel. These findings can provide theoretical support for the design of tunnels crossing creep-slip and stick-slip faults.