The dynamic responses of subgrade bed layers are the key factors affecting the service performance of a heavy‐haul railway. A 3D train‐track‐subgrade interaction finite element (FE) model was constructed using the ABAQUS code, where different vertical irregular track spectra were simulated by modifying the vertical node coordinates of the FE mesh of the rail. Then, the dynamic stresses in the subgrade bed layers subjected to heavy‐haul trains were studied in detail. The results showed the following: (1) the transverse distribution of the dynamic stress transformed from a bimodal pattern to a unimodal pattern with increasing depth; (2) the pass of adjacent bogies of adjacent carriages can be simplified once loaded on the subgrade since the dynamic stresses are maintained around the peak value during the pass of the adjacent bogies; (3) the dynamic stress at the bottom of the subgrade bed surface layer was more sensitive to the train axle load compared with that at the subgrade surface because the dynamic stresses induced by the two rails were gradually overlaid with increasing depth; (4) the maximum dynamic stress at the subgrade bed bottom was reduced by approximately 70% compared with that at the subgrade surface; (5) the vertical track irregularities intensified the vertical excitation between the train vehicle wheels and rails, and the maximum dynamic stress at the subgrade surface under the action of the irregular heavy‐haul track spectrum increased by 23% compared with the smooth rail condition; and (6) the possible maximum dynamic stress (σ_dm) at the subgrade surface under the action of irregular track spectra can be predicted using the triple standard deviation principle of a normally distributed random variable, i.e., σ_dm = μ + 3σ (where μ and σ are the expectation and standard deviation of σ_dm, respectively).

With the popularized application of micropile for slope reinforcement, there are many kinds of pile section form, such as single steel bar, multisteel bar, and tube. In order to obtain the theoretical calculation model of horizontal capacity of micropiles with different types, numerical simulations for two common conditions, namely, soil‐soil and soil‐rock, were carried out to study micropile reinforced slope’s deformation and failure regulation. The dominated indicator for micropile horizontal capacity was acquired. By using the equivalence principle, four kinds of micropiles’ capacity calculation model were deduced. Then, FEM numerical analysis was used for validation. The results show that the flexural capacity of micropile is a critical and dominated indicator which should be considered seriously in slope reinforcement design. When the ultimate flexural capacity is reached, the displacement of micropile reinforced slope will increase rapidly. The primary factor affecting the flexural capacity of micropiles is pile diameter, followed by section reinforcement ratio. The calculation results of the calculation model of flexural capacity proposed in this paper are close to those of numerical ones and are safer. The calculation model can provide reference for micropile selecting and slope reinforcement design, especially for quick design of emergency engineering.

In mountainous areas, high-filled subgrade often experiences significant post-construction settlement. Prematurely paving the cement-stabilized gravel layer on an unstable subgrade can easily lead to subsequent cracking. To accurately predict the settlement of high-filled subgrade and determine the appropriate timing for paving the cement-stabilized layer, this study proposes a subgrade settlement prediction method combining an Artificial Neural Network (ANN) with a Genetic Algorithm (GA). Using monitoring data from a high-filled subgrade on a highway in Hunan Province, China, a GA–ANN model was established to predict settlement curves. The predicted data from the GA–ANN model were compared with measured data and ANN predictions to validate the advantages of using GA–ANN for subgrade settlement prediction. The results indicate that the GA–ANN model significantly outperforms the ANN model due to GA’s ability to provide more reasonable weight biases for ANN through global search optimization. Predictions of settlement data beyond 50 days using both ANN and GA–ANN showed that the GA–ANN prediction curve closely matched the measured curve, with a basic deviation within ±3 mm. In contrast, ANN’s prediction error gradually increased to over 5 mm as the observation time increased, with predicted values lower than measured values, leading to an overly optimistic estimation of early settlement convergence. Based on the predicted data and settlement standards, the estimated timing for laying the stabilized layer was determined. During the laying process, no cracking was observed in the stabilized layer. The project has been in operation for six months, with the road surface in good condition. This study provides a valuable reference for the laying of stabilized layers on similar high-filled and ultra-high-filled subgrades.

With the enhancement of transportation speed and axle load, dynamic response of subgrade increases significantly. In order to improve the calculation accuracy of subgrade response under complex stress state, it is necessary to use dynamic indicators instead of static indicators in calculative process. For the sake of investigating the influence factor of dynamic resilient modulus of subgrade silty clay in Eastern Hunan, resilient modulus tests were carried out by conducting repeated load tri-axial tests. Based on available model, an improved resilient modulus prediction model considering four parameters was proposed by introducing k4. Corresponding accurate consistent tangent stiffness matrix was derived. Afterward, the improved model was implemented into finite element method software and verification work was put forward both on single element and pavement-subgrade structure. Finally, calculated results were compared with in-site measured results. Study achievements demonstrate that the improved model exhibits a higher precision and efficiency on single element because k4 can better adjust the affecting proportion of octahedral shear stress. When applied to analysis on pavement-subgrade structure, the improved model can reflect subgrade resilient modulus distribution and evolution more factually. In addition, numerical calculated result nearly coincides with measured results, which shows the application value of the improved model.