Mountain tunnels in western China are mostly located in regions with high seismic intensity and complex geological and topographic conditions, where active faults are densely distributed. Many tunnels are constructed in near-fault strong earthquake zones. For example, in the ongoing construction of the Sichuan-Tibet Railway, tunnels account for as much as 70% of the total route, and 54 active faults are distributed along and near the alignment. Numerous seismic investigations have shown that near-fault ground motions have a significant destructive impact on tunnel structures. Notable cases include the 1999 Chi-Chi earthquake in Taiwan, the 2008 Wenchuan earthquake in Sichuan, the 2016 Kumamoto earthquake in Japan, and the 2022 Menyuan earthquake in Qinghai, all of which resulted in severe tunnel damage.

The long-period velocity pulses commonly present in near-fault ground motions are considered key factors contributing to structural damage. However, due to the limited availability of strong earthquake observation records, especially near engineered structures, near-fault pulse-type ground motion records remain scarce. As a result, many existing studies have relied on synthetic ground motion methods to generate seismic time histories for analysis. Currently, research on the seismic response of mountain tunnels under near-fault pulse-type ground motions remains limited, particularly concerning the mechanisms through which pulse characteristics influence tunnel seismic responses, which are still not well understood.

On the other hand, topographic amplification effects induced by mountain terrain also significantly affect tunnel seismic responses. Studies have shown that mountain topography can amplify peak ground velocity （PGV） due to multiple reflections of seismic waves on the surface, leading to localized wave focusing effects and significant spatial variability within the mountain. However, most existing studies on mountain tunnel seismic responses commonly adopt full-space or half-space site models, neglecting topographic effects. Additionally, variations in the incident wave angle can alter the propagation path and energy distribution of seismic waves, further increasing the spatial variability of seismic damage.

This study comprehensively considers the coupled effects of near-fault ground motion characteristics and topographic amplification and proposes a systematic research framework. First, artificial synthesis techniques are employed to generate near-fault pulse-type ground motions. The high-frequency components of two representative recorded ground motions serve as the basis, while the low-frequency components are fitted using an equivalent pulse model. By adjusting pulse parameters, ground motions with different spectral characteristics are synthesized. Second, based on an indirect boundary element method-finite element method（IBEM-FEM）coupled method, a cross-scale efficient analysis was achieved, bridging the amplification effects in kilometer-scale complex sites to the dynamic damage of tunnel cross-sections at the centimeter scale, thereby revealing underlying seismic response patterns. Finally, taking pulse amplitude, pulse period, and pulse number as key parameters, the influence of near-fault pulse-type ground motions with different incident angles on tunnel seismic responses is examined, revealing the damage and failure patterns of tunnel linings.

The results indicate that tunnel lining damage severity increases with pulse amplitude. Compared with directivity effects, fling-step effects are more sensitive to changes in pulse amplitude, and oblique incidence combined with high-amplitude pulses can cause more severe structural damage. Tunnel damage severity decreases as pulse period increases, with directivity effects being more sensitive to pulse period variations. Under oblique incidence conditions, tunnel structural damage severity increases with the number of pulses.

For seismic design of mountain tunnels in near-fault regions, the combined impact of near-fault pulse effects and topographic amplification should be carefully considered, and tunnel seismic strategies should be optimized. Firstly, the locations of seismic wave focusing and amplification within the mountain should be identified and avoided to reduce the adverse effects of localized seismic motion amplification on tunnel structures. Secondly, for high-amplitude pulse effects, it is recommended to strengthen the seismic design of tunnel linings, particularly in areas where fling-step effects are prominent, to mitigate the damage caused by extreme ground motions. Additionally, during the design process, the influence of seismic wave incidence angle should be fully considered, and reinforcement measures for tunnel linings should be implemented accordingly to enhance tunnel seismic performance under complex wave propagation conditions.