The widely exposed granite bedrock contains mafic dike clusters, fractures, and faults. Figure 3 illustrates a detailed structural map of these features in the studied outcrop. The fracture network shows two dominant trends: ENE-WSW (or E‐W) and NNE‐SSW (or N‐S; Figures 3, 4). NE-SW striking fractures, exhibiting good continuity, cross-cut the two dominant fracture sets and enhance the connectivity of the grid-like network. The mafic dike cluster region, west of the main fault, exhibits a grid-like fracture network, with fracture orientations systematically varying with proximity to the fault. The main fault, which extends over 250 m, both cuts through and follows the boundaries of highly deformed NE-striking dikes, and is characterized by 5–10 cm thick gouge zones and brecciation within and along the margins of these dikes. The main fault gradually diminishes in intensity toward the north. Smaller-scale faults, trending NNE and E–W, occur within and along the margins of the dikes, with fault strikes that are parallel or sub-parallel to the dike orientation, particularly at points ofbending or branching. Hydrothermal alteration, including calcite vein formation, is widespread along extensional fractures and faults, particularly within fault zones and tip damage zones.

Dike trajectories are controlled by mechanical interactions between shear stresses and pre-existing fractures, with host rock heterogeneity and magmatic overpressure further influencing propagation dynamics. At dike tips, tensile fractures transition to shear fractures as magmatic pressure and host rock heterogeneity rotate principal stress orientations (Gudmundsson, 2002; Dering et al., 2019). This stress reconfiguration drives the development of en-echelon fracture arrays and promotes fault reactivation through localized strain accumulation. In the study area, three distinct dike orientations reflect these stress-fracture interactions (Figures 3, 4): ENE-striking dikes exhibit sharp, planar boundaries and extensive lateral continuity, terminating at or intruding along NNE-oriented fractures (Figure 1B) or developing trailing geometries (Figure 1C). NNE-striking dikes display limited continuity, forming sinuous (zigzag) geometries where they interact with ENE-striking dikes. NE-striking dikes maintain lateral continuity despite irregular boundaries and variable thickness, occasionally exploiting pre-existing NNE- or ENE-striking fractures as pathways.

The direction of dike extension typically aligns with the σ₃ orientation. However, in regions with complex geometries caused by passive intrusion, such as dike bending, branching, or irregular terminations, this relationship becomes unclear. The three dominant dike orientations (Figure 4) indicate variable propagation paths governed by local rotations of stress field or pre-existing fracture networks, complicating the identification of the σ₃ direction. To address this uncertainty, geometric restoration methods were applied to resolve complex stress fields and reconstruct stress regimes associated with non-linear fracture patterns.

In this study, dike boundaries were geometrically restored, and their relationships with fractures were analyzed (Figure 5). Dikes were first extracted from the bedrock (Figure 5A), with kink points along dike boundaries identified as piercing points for restoration (Figure 5B). The shortest path connecting these points revealed a dilation direction of approximately 332° (σ_Hmin: 332°, σ_Hmax: 062°; Figures 5C,D). Although the shortest-path reconstruction indicated a consistent dilation direction of approximately 332°, slight mismatches along dike boundaries remained. To resolve these mismatches, the northernmost fracture was adjusted by a 3° counterclockwise rotation, which aligned the fractures into a pre-intrusion configuration. This adjustment suggests that the mismatches resulted from oblique dike intrusion or irregular fracture growth along dike margins. Additionally, restricted fracture propagation at dike tips could induce localized stress variations, leading to block rotation within the dike body. The main dike (deformed dike) extends over 250 m with irregular boundaries and complex branching, yet maintains overall lateral continuity representing the primary extensional direction (Figure 3). We interpret this 332° dilation as reflecting the local minimum principal stress (σ₃) orientation during emplacement, indicating NW–SE extension and σ_Hmax of ∼062°. However, dilation represents strain and does not necessarily coincide with stress orientation. This interpretation pertains to the dominant NE-trending extensional system, whereas dikes that intruded passively along pre-existing ENE and NNE fractures reflect structural inheritance rather than the prevailing ambient stress conditions.

In the study area, dikes are observed intruding into orthogonal fracture networks, with representative examples in Figure 6. Irregular NE- and ENE-striking dikes exhibit blunt-ended tips and fault-controlled geometries, indicating passive intrusion where magma follows pre-existing fractures and is deflected by structural barriers (Figure 6A). Some thin dikes (centimeter-scale) with narrowing tips form independently and terminate at N-S-striking faults. The dike geometry, including offsets and thickness variations, implies structurally controlled emplacement. The N-S-striking faults create apparent left-lateral displacement along the western fault segment and apparent right-lateral displacement along the eastern fault. This displacement is termed “apparent” because it results not from post-emplacement fault slip but from the inability of intruding magma to propagate across the pre-existing discontinuity. Instead, magma was deflected along the fault plane, creating discontinuous dike segments that appear offset in map view. The asymmetric thickness and distribution variations across the fault confirm syn-emplacement structural control rather than post-intrusion faulting.

A dike approximately 5 cm wide confined between two faults displays a zigzag geometry (Figure 6B). The thickness of these zigzag-patterned dikes varies with fracture orientation, and some partially intrude N-S-striking fractures (Figure 6B). These features indicate passive dike emplacement guided by pre-existing fractures. Irregular dike propagation along fractures results in displacement mismatches, causing terminations or deflections at fracture orientation changes (Figures 3, 6B). Dikes terminate at the fractures, with some partially intruding along these fractures rather than being fully offset (Figure 6C). The NE σ_Hmax (062°) imposed dextral shear along the fractures and controlled NE-striking dike alignment, consistent with a stable stress regime during emplacement.

En-echelon E-W-striking dikes formed as overlapping or stepping sub-parallel fractures under an E-W σ_Hmax, commonly observed in extensional and transtensional settings (Figures 3, 7). In fault growth models, such fractures develop in bedrock during early fault evolution or as a response to distributed shear strain (Mollema and Antonellini, 1999; Peacock, 2001). The mapped en-echelon dikes extend to approximately 6 m, with width of 0.3 to 0.7 m and spacing of 2.4–3.1 m. Their widths are greatest at the midsection and gradually taper toward the edges.

The E-W-striking en-echelon fractures could originate as extensional features under an E-W σ_Hmax stress which may also explain their association with right-lateral shear deformation (Figure 7). During dike emplacement, the regional stress regime shifted from an earlier E-W to NE σ_Hmax (Figure 5), indicating a significant reorientation of the principal stress axis. This shift is supported by sigmoidal (curved) dikes within shear zones, which resemble tension gashes and indicate NE σ_Hmax (Figure 7). Additionally, NE-striking veins and fractures are confined between dikes, without extending beyond their boundaries. The geometry of connecting fractures suggests sinistral shear displacement. NE-striking fractures between dikes contain vein infill, indicating post-emplacement hydrothermal activity.

Significantly, subsidiary veins confined between E-W dikes exhibit geometric characteristics identical to linking damage zones observed in reactivated E-W sinistral fault systems elsewhere in the study area. This indicates a three-stage evolution: first, E-W σ_Hmax generated initial en-echelon fractures, second, NE σ_Hmax controlled dike intrusion along these fractures, producing curved geometries at terminations under normal fault conditions; and third, continued NE σ_Hmax led to post-emplacement sinistral deformation under strike-slip fault conditions, creating linking damage zone characteristics between dike segments.

Calcite-dominated veins preferentially filled pre-existing fractures aligned with dike orientations and clustered along dike margins and interiors, where hydrothermal fluids intensely altered the dike (Figures 8A,B). Vein formation occurred after dike cooling, as indicated by cross-cutting relationships. Late-stage hydrothermal activity produced a stockwork vein network and brecciated dike interiors through selective dissolution of mechanically weakened zones (Figure 8B).

Stress concentrations at dike-bedrock interfaces (driven by stiffness contrasts and mechanical anisotropy) localized fracture propagation and vein mineralization (d’Alessio and Martel, 2005; Gwon and Kim, 2016). ENE-striking sinistral faults and associated NE-SW secondary fractures also contain veins (Figures 8C,D), indicating that fault reactivation occurred after dike emplacement.

The fracture system in the study area is dominated by two principal orientations. Several ENE-striking fractures are reactivated as sinistral faults under NE horizontal stress (Figures 3, 8C, 9). One of the sinistral faults in the study area is associated with an 8 m-wide tip damage zone containing densely clustered secondary fractures (averaging 10 m in length) that trend 045° with steep dips (Figure 8D).

Detailed mapping on an excavation surface (Figure 9A) reveals ENE-striking sinistral faults with left-stepping geometries and vein-filled linking damage zones. Secondary fractures concentrate along fault segments and within linkage zones, where fracture density peaks. In the southeastern part of Figure 9A, a sinistral fault (thick red lines) crosscuts a dike, forming breccia and branching into multiple segments. Fracture density analysis indicates higher values east of the NE-striking dike (Domain 1) compared to the west (Domain 2; Figure 9A).

To investigate the structural interactions of DF² system, we employed a topological analysis in the SE sector of the study area (Figures 3, 9). Digitized fracture traces were analyzed using QGIS’s NetworkGT toolkit to calculate two metrics (P₂₁ and C B ).

During dike intrusion, magma-driven stresses (e.g., overpressure and shear coupling along fractures) reorient the local stress field. These stress perturbations can trigger slip along pre-existing fractures or generate new tensile fractures parallel to the NE σ_Hmax (062° ± 10°), which dominated during the main phase of dike emplacement (Figures 3, 10A). These fractures and associated normal faults act as preferential magma pathways, contributing to fault system maturation through cyclical damage accumulation (Muirhead et al., 2016; Currenti et al., 2019). In our study area, these fractures and associated normal faults subsequently served as preferential pathways for magma migration, promoting cyclical damage accumulation and the progressive maturation of the fault zones. This process is evidenced by sigmoidal dike geometries and overprinting slickenlines (Figures 7, 10E,F). Repeated intrusions convert fault zones into complex networks, generating secondary fractures, subsidiary faults, and branching geometries. Mature fault systems accommodate dikes with complex geometries, contrasting with simpler geometries in undeformed bedrock (Spacapan et al., 2016).

The NE-striking dikes form laterally continuous networks (>250 m) displaying asymmetric and fault-parallel branching geometries (Figures 3, 10). Dense fracturing occurs within dike interiors and along margins that run parallel to adjacent faults, merging into ∼8–10 m wide deformation zones surrounding those faults. These zones contrast sharply with narrow fault gouge zones (∼5 cm) along slip surfaces. Faults develop along dike boundaries and interiors, displacing dike margins and internal veins (Figures 3, 10A,C). Blunt-ended dike tips and their injection into fractures aligned with main faults (Figures 10B,D) indicate that emplacement was controlled by pre-existing or simultaneous tensile fractures. Slickenlines record normal-oblique slip (plunge 33°, 46°) and strike-slip (plunge 1°) movements (Figures 10E,F). Superimposed slickenlines indicate post-intrusion clockwise stress rotation (Δ σ_Hmax = ∼20°) and multi-phase reactivation under normal to dextral fault movement (Gwon and Kim, 2016).

The NE-striking main fault dips northwest and accommodates a dike cluster in its hanging wall, while fracture patterns differ between the hanging wall and footwall (Figure 3). This suggests the main fault may act as the boundary of a local graben during dike emplacement (Rahimi et al., 2025). Geometric similarity between dikes and main faults, including branching patterns and repeated fault reactivations along dikes, indicates that pre-existing normal faults provided pathways for dike emplacement. Fault nucleation within dike interiors and shear-coupled reactivations further support this process.

Dikes initially used tensile fractures (Mode I) under a NE-striking σ_Hmax, as shown by fracture arrays parallel to the dike. Subsequent fault movements, transitioning from normal to strike-slip, as recorded by striae overprinting (Figures 10E,F), reflect a clockwise rotation of the post-emplacement stress field from a normal to a strike-slip regime. Repeated reactivation along dike interiors and boundaries demonstrates their role as inherent structural weaknesses, facilitating strain localization during tectonic events.

These observations demonstrate that dike-fault interactions are governed by syn-emplacement stage coupling of magma-driven fracturing and tectonic shear, rather than separate tectonic episodes.

Forceful intrusions generate new fractures by exceeding the tensile strength of intact rock, whereas passive intrusions utilize pre-existing fractures under moderate magmatic pressure. These mechanisms often coexist, reflecting the interplay between magmatic pressure and structural fracture complexity of the bedrock (Pollard, 1987; Lister and Kerr, 1991; Martínez-Poza et al., 2014). The emplacement of dikes in the study area occurs through two primary emplacement mechanisms.

The ENE- and ENE-striking fractures form a grid-like pattern with dikes intruding along fractures (Figures 1B, 3, 6C) or partially penetrating them (Figures 1C, 3, 6A). Zigzag geometries indicative of passive intrusion are observed at various scales (Figures 3, 5, 6B). Restoration of these geometries suggests a σ_Hmax orientation of 062° aligned with forceful dike planes (Figures 3, 6A, 10B). Dike intrusions associated with shear deformation form en-echelon structures (Figure 7) and reactivate N-S-striking fractures (Figure 6). NE-striking dikes follow primary extension, generating fault parallel or sub-parallel fractures during forceful emplacement. Post-dike emplacement faulting involved hydrofracturing (Figure 9).

A hybrid model integrating passive and forceful intrusion mechanisms is proposed based on our observations (Figure 11). Throughout the syn-emplacement stage, dikes employ a dual mechanism: exploiting pre-existing fractures in the basement rock and propagating through intact bedrock via magmatic overpressure. Post-emplacement ENE-striking sinistral movement occurs under a rotated stress field where the NE σ_Hmax becomes dominant (Figures 7, 8C, 9A). This motion is facilitated by inversion of the σ₁ and σ₂ axes during dike emplacement, which also triggers hydrothermal alteration along reactivated fault zones. Stress orientations inferred here assume regional homogeneity, but local variations could arise from complex magma chamber geometries (e.g., multi-chamber systems; Hazarika et al., 2024). As our study area lacks exposed chambers, this remains speculative. Future work should integrate geophysical constraints to resolve such uncertainties.

Hybrid emplacement appears broadly applicable, depending on three interdependent controls. (1) Depth–rheology: At shallow levels (<∼5 km) low mean stress promotes forceful, mode-I dike (Brace and Kohlstedt, 1980; Rybacki et al., 2021). Within the brittle–ductile transition (5–15 km) both tensile failure and structural reactivation coexist, yielding true hybrids (this study). Below ∼15 km, high-temperature ductile flow confines magma to pre-existing anisotropies (Klepeis et al., 2022; Chatterjee et al., 2024). (2) Host lithology: sedimentary successions generate sill-dike networks along bedding (Spacapan et al., 2016; Magee et al., 2016), whereas high-grade gneiss directs magma along foliations (He et al., 2020; Zhang et al., 2024). (3) Magma rheology: low viscosities (∼10²–10⁴ Pa s) favour pervasive fracture infiltration, intermediate values (∼10⁴–10⁶ Pa s) optimize hybrid behavior, and high viscosities (>∼10⁶ Pa s) bias intrusion toward fault-controlled pathways (Giordano et al., 2008).

These results from the study area demonstrate that dike emplacement systems result from the coupled effects of structural inheritance and magmatic forcing. Pre-existing fractures facilitate passive magma intrusion, while magma overpressure generates new fractures via forceful intrusion. This dual mechanism, where structural inheritance and structural generation operate throughout the syn-emplacement stage, highlights the need for integrated models to resolve crustal deformation in pre-existing basement rocks.