ANIDIS - L'ingegneria Sismica in Italia, ANIDIS XIX & ASSISi XVII - 2022

Dimensione del carattere:  Piccola  Media  Grande

Machine-learning-enhanced variable-angle truss model to predict the shear capacity of RC elements with transverse reinforcement

Dario De Domenico, Giuseppe Quaranta, Qingcong Zeng, Giorgio Monti

Ultima modifica: 2022-08-03


The prediction of the shear capacity of reinforced concrete (RC) elements with transverse reinforcement is a critical topic to which a huge amount of researches have been devoted over the last four decades. This is motivated by the fact that several existing RC structures are provided with transverse reinforcement much lower than that recommended in current design codes, and thus often exhibit a shear-dominated failure. In order to perform an accurate vulnerability assessment of such structures, it is of utmost importance the development of reliable, unbiased, and precise numerical formulations capable of predicting the actual shear strength of RC elements with stirrups. Some formulations proposed in the past years within international codes were found to be overly conservative compared to experimental findings and are often characterized by excessive dispersion. On the other hand, the use of machine learning techniques has been also exploited recently to obtain more accurate pure data-driven predictions of the shear capacity of RC members. 

Instead of adopting a pure data-driven approach to derive new empirical capacity equations as proposed by ongoin studies, the strategy devised in this contribution employs artificial intelligence tools for ehancing one of the most popular mechanical models adopted in international codes for the shear capacity prediction of RC elements with stirrups, i.e., the variable-angle truss model. More specifically, genetic programming is used to calibrate two coefficients ruling the concrete contribution in such capacity model, so as to increase its accuracy. In this way, the mechanical basis of the resisting mechanism is preserved, but the correctness of the final predictions is improved thanks to such hybridization. Effectiveness and potentials of the proposed formulation are demonstrated by comparison with experimental results collected from a large database including RC beams and columns with solid and hollow cross-sections failing in shear under monotonic or cyclic loading. A comparative analysis is also made for a large set of expressions from reference building codes to show that the proposed unified shear capacity equation leads to more accurate outcomes and, ultimately, to prove that it is suitable for practical design applications.

è richiesta l'iscrizione al convegno per poter visualizzare gli interventi.