Numerical Investigation and Machine Learning Prediction of Dynamic Increase Factor for Reinforced Concrete Beams Under Various Loading Rates

The mechanical behavior of reinforced concrete (RC) beams is significantly affected by the loading rate. High loading rates influence not only the load-carrying capacity and deformation capacity of RC beams but also their failure modes, which further alter the seismic performance of RC beams. On the basis of previous experimental tests, extensive dynamic numerical simulations of RC beams under a wide range of loading rates are conducted using ABAQUS 2021 to investigate the effects of loading rate, shear span ratio and reinforcement ratio on the mechanical performance of RC beams. Numerical results indicate that elevated loading rates increase the dynamic load-carrying capacity and energy dissipation of the RC beams. The numerical framework predicts a transition in failure mode from brittle shear to ductile flexure under extreme loading rates. However, this transition is highly sensitive to the perfect-bond assumption, indicating that dynamic bond–slip behaviors should be incorporated in future assessments. A database of 213 records was compiled, comprising 162 full-factorial numerical cases and 51 experimental tests. Machine learning surrogate models were trained on the numerical dataset using a 5-fold validation strategy to decouple parametric influences. The surrogate model analysis demonstrates that the loading rate is the primary parameter governing the dynamic enhancement. Experimental data were subsequently used for qualitative and quantitative validation to define the physical applicability domain of the numerical models.