Reconciliation of computational fluid dynamics and observations in complex terrain through conditional resampling. Issue 195 (December 2019)

• Record Type: Journal Article
• Authors: Nakao, Keisuke; Hattori, Yasuo
• Is Part Of: Journal of wind engineering and industrial aerodynamics. Issue 195(2019)
• Publication Date: 2019-12
• Subjects: Large-eddy simulation -- Complex terrain -- Turbulence statistics; Wind-pressure -- Periodicals; Buildings -- Aerodynamics -- Periodicals; Pression du vent -- Périodiques; Constructions -- Aérodynamique -- Périodiques; Buildings -- Aerodynamics; Wind-pressure; Periodicals
• DOI: 10.1016/j.jweia.2019.103970
• Languages: English
• ISSNs: 0167-6105

Abstract

Abstract: The 10-min statistics of the turbulence observed on masts often vary with wind speed. On the other hand, the wind engineering computational fluid dynamics (CFD) analysis often carried out under adiabatic conditions assumes Reynolds number independence. Provided the prescribed upwind condition of CFD corresponds to the ensemble of strong wind, conditional sampling and averaging the observation may be a feasible process to correlate CFD results with the observation. Using comprehensive observation data obtained from complex terrains and operating a suite of large-eddy simulations with O (10 m) horizontal grid resolution, we demonstrate how observation with ensemble averages and conditional samplings can be reconciled with CFD. We find that selecting a high wind speed (approximately 10 m/s at 60 m above ground level) and taking the average of ten sets of 10-min data (corresponding to a sample of 6000 points at 1 Hz) yields better consistency with CFD results. We demonstrate that conditional sampling is more effective than increasing the ensemble size. This tendency is further confirmed when the ensemble size is evaluated by how many times eddies pass through the observation mast. Highlights: The observations and large-eddy simulations over complex terrains were carried out. How observation with conditional samplings and ensemble averages can be reconciled with CFD was investigated. Selecting a high speed and taking the average of ten sets of 10-minute observation data yield better consistency with the CFD.