Akademik Veri Yönetim Sistemi
Structures, cilt.87, 2026 (SCI-Expanded, Scopus)
AbstractAccurate modelling of the Textile Reinforced Mortar (TRM)–masonry interface is essential for predicting the effectiveness of TRM strengthening, yet perfect-bond (tie) constraints suppress slip/separation and can overestimate stiffness and capacity. This study presents a numerical framework for implementing generalized TRM–masonry bond–slip laws in cohesive traction–separation form. Using a database of 48 pull-off tests obtained from a previously conducted experimental campaign, multivariable linear regression models were developed for key bond-law descriptors (peak stress, residual stress where applicable, characteristic slip values, ultimate slip, and fracture energy) as functions of mortar compressive strength, strip width, bonded length, anchorage, and masonry unit type (hollow clay brick, solid clay brick, and autoclaved aerated concrete). Experimental results indicate a trilinear response with residual stress for clay brick substrates, whereas autoclaved aerated concrete exhibits an inherently bilinear response decaying to zero. Because the standard Abaqus cohesive implementation is bilinear, an energy-based transformation is proposed to map trilinear laws to an equivalent bilinear triangular traction–separation relation while preserving fracture energy. The resulting interface definitions were implemented in Abaqus and validated against the 48 load–displacement curves. The energy-equivalent approach reproduces peak response and global energy dissipation with close agreement, with the largest deviations in stiffness-related predictions for brick substrates. Wall-scale diagonal compression simulations further confirm good performance for hollow-brick walls, while discrepancies for solid clay-brick walls highlight the need for substrate-specific tuning when scaling up. Overall, the proposed approach provides a direct and transferable route to incorporate experimentally informed TRM–masonry interface laws into cohesive-zone finite element models.