On the use of ensemble averaging techniques to accelerate the Uncertainty Quantification of CFD predictions in wind engineering

In this work we focus on reducing the wall clock time required to compute statistical estimators of highly chaotic incompressible flows high performance computing systems. Our approach consists replacing a single long-term simulation by an ensemble multiple independent realizations, which are run in parallel. A failure probability convergence criteria must be satisfied estimator interest assess convergence. The error analysis leads identification two contributions: initialization bias and error. We propose systematically detect burn-in needed order minimize bias, as well techniques choose effective keep under control. This is accompanied strategies reduce cost. proposed method assessed application prediction drag force over rise buildings specifically CAARC building, relevant benchmark for wind engineering community. • Development framework analyze computational fluid dynamics Solution systems characterized ergodic nonergodic uncertainties. Definition accelerate solution CFD Proposal initial condition strategy ensuring independence realizations. Statistical quantities (drag force, base moment, pressure field) using raw central moments.