PSI - Issue 78

Enes Krasniqi et al. / Procedia Structural Integrity 78 (2026) 261–268

267

between the plateau phases observed experimentally and in the simulation. The simulation did not attain convergence at large displacements. The crack pattern was simulated with good accuracy, both concerning the tensile cracks and the conical breakout failure cracks. Overall, the excellent agreement in crack orientation, failure surface geometry, and the global failure mechanisms validates the numerical approach employed. Nonetheless, future improvements could include more sophisticated modelling of aggregate bridging, distributed micro-cracking, and post-cracking residual friction, to further narrow the gap between simulated and experimental post-peak behaviour. The mechanical nut substantially altered the failure mode, changing the abrupt failure driven by interface debonding of specimen TR01b into a concrete breakout failure for nutted specimens TR02a and TR02b. The embedment length proved being a fundamental parameter to enhance ductility. As shown by further testing, to proper longer embedment length is associated tensile failure of the rebars, maximising ductility, ensuring a proper seismic behaviour of the joint.

Figure 6. TR01b: Load–displacement, damage and crack pattern comparison between experimental results and ATENA finite element model.

Figure 7. TR02a: Load–displacement, damage and crack pattern comparison between experimental results and ATENA finite element model.

Figure 8. TR02b: Load–displacement, damage and crack pattern comparison between experimental results and ATENA finite element model.