مطالعه عددی رفتار محل اتصال در پل های پیش‌ساخته با بتن فوق توانمند

This study presents a comprehensive numerical investigation into the shear behavior of dry joints in precast segmental bridges constructed with ultra-high-performance concrete (UHPC). Given the critical role of joint performance in ensuring the continuity and integrity of precast bridge structures, an advanced three-dimensional finite element (FE) model was developed using the Abaqus software platform. The model incorporates the Concrete Damaged Plasticity (CDP) constitutive law and is calibrated against experimental data to ensure its validity, showing excellent agreement with test results (less than 5% discrepancy in peak load prediction). In the proposed modeling framework, UHPC segments, shear keys, and loading blocks are represented by three-dimensional solid elements, while the joint interface is captured through a surface-to-surface contact formulation that allows separation in the normal direction and frictional sliding in shear. Material parameters for the CDP model in both compression and tension are selected based on available UHPC test data and then fine-tuned so that the numerical model reproduces the initial stiffness, peak shear capacity, and post-peak softening response of a benchmark push-off test. This calibration procedure ensures that the simulated load–displacement curves and crack patterns closely follow the experimental observations, thereby providing a robust and reliable numerical basis for the subsequent parametric study on the influence of confining stress and shear-key geometry. A parametric study was subsequently conducted to evaluate the effects of three governing parameters: (1) number of shear keys (1, 2, and 3), (2) shear key depth-to-height ratio (D/H = 0.5, 0.6, 0.7) and (3) magnitude of lateral confining stress (σc= 3, 5, 10 MPa). A total of 12 models were analyzed and force–displacement responses were extracted to compute key performance indices: peak load (Pmax) , displacement at peak, initial stiffness, and residual force. Results revealed that increasing confining stress leads to higher Pmax initial stiffness, particularly in single-key specimens. Likewise, an increase in the number of shear keys significantly improves both load-carrying capacity and post-peak stability. Key depth also enhances initial stiffness and shear resistance, but excessive depth may induce more brittle post-peak behavior, especially in single-key joints. All samples demonstrated a characteristic three-stage response: a steep initial ascent, followed by a distinct peak, and then a softening phase that converged to a residual plateau. The observed damage initiated at the upper shear key and progressed downward, consistent across multi-key configurations. Interestingly, peak loads occurred at small displacements (0.48–0.70 mm), highlighting the localized and early activation of joint resistance. Increases in capacity were mostly attributed to steeper ascending branches, rather than larger peak displacements. To assess the accuracy of existing analytical models, four shear strength prediction equations were evaluated: AASHTO, Voo et al., Gopal et al. and Oettel & Empelmann. The predicted capacities were compared with FE results for all 12 specimens. Among them, the AASHTO model showed the highest consistency, with a mean Vu/Vu,fe ratio of 0.73. The Oettel model yielded acceptable results (mean ratio: 0.58), while the Voo and Gopal models significantly underpredicted shear capacity, with mean ratios of 0.35 and 0.36, respectively. In conclusion, the proposed FE framework offers a reliable tool for analyzing UHPC joints, and the findings suggest that confining stress and key configuration play critical roles in joint performance. Additionally, current code-based equations may require refinement to adequately capture the shear behavior of UHPC dry joints in precast segmental bridge systems. Based on the FE database, a simplified design-oriented equation is also proposed to estimate the shear capacity of UHPC dry joints as a function of confining stress, shear-key geometry, and number of keys, showing good agreement with the numerical results over the investigated range.