In the present paper is described an alternative algorithm for the simulation of dynamic response of tall building subject to the turbulent wind load, the present technique is part of the approach called Performance Based Wind Engineering which is traditionally used in combination to the Brute Force approach (BF) for solving problem characterized by randomness, which is been utilized to describe the stochastic nature of the drag coefficient CD. The present technique utilizes a statistical tool based on the Robbins-Monro theorem used as nonlinear root finding of the generic function contaminated by an error, the method uses a simple recursive formula which converges to an approximate solution of the average value ̅ ∗ of the outgoing distribution, moreover a new approach aimed to the evaluation of the standard deviation value of the output distribution was applied in the present work. Verification of the proposed algorithm was performed through Monte-Carlo method, after preselecting three different probability distribution typologies, each characterized by the mean μCD and the coefficient of variation σCD/μCD. A simplified structure characterized by a single DOF is been employed as testing structure for validation purposes.
Stochastic Approach for the dynamic performance analysis of tall buildings subject to turbulent wind load / Giaccu, Gian Felice; Scintu, Luca; Barbiellini, Bernardo. - (2015). (Intervento presentato al convegno ICWE14-14th International Conference on Wind Engineering tenutosi a Porto Alegre, Brazil. nel June 21-26 2015).
Stochastic Approach for the dynamic performance analysis of tall buildings subject to turbulent wind load
Gian Felice Giaccu
;
2015-01-01
Abstract
In the present paper is described an alternative algorithm for the simulation of dynamic response of tall building subject to the turbulent wind load, the present technique is part of the approach called Performance Based Wind Engineering which is traditionally used in combination to the Brute Force approach (BF) for solving problem characterized by randomness, which is been utilized to describe the stochastic nature of the drag coefficient CD. The present technique utilizes a statistical tool based on the Robbins-Monro theorem used as nonlinear root finding of the generic function contaminated by an error, the method uses a simple recursive formula which converges to an approximate solution of the average value ̅ ∗ of the outgoing distribution, moreover a new approach aimed to the evaluation of the standard deviation value of the output distribution was applied in the present work. Verification of the proposed algorithm was performed through Monte-Carlo method, after preselecting three different probability distribution typologies, each characterized by the mean μCD and the coefficient of variation σCD/μCD. A simplified structure characterized by a single DOF is been employed as testing structure for validation purposes.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.