Abstract—The flow field over a three dimensional pole barn characterized by a cylindrical roof has been numerically investigated. Wind pressure and viscous loads acting on the agricultural building have been analyzed for several incoming wind directions, so as to evaluate the most critical load condition on the structure. A constant wind velocity profile, based on the maximum reference wind speed in the building site (peak gust speed worked out for 50 years return period) and on the local roughness coefficient, has been simulated. In order to contemplate also the hazard due to potential air wedging between the stored hay and the lower part of the ceiling, the effect of a partial filling of the barn has been investigated. The distribution of wind-induced loads on the structure have been determined, allowing a numerical quantification of the effect of wind direction on the induced stresses acting on a hemicylindrical roof. Keywords—CFD, wind, building, hemicylindrical roof. I. INTRODUCTION AND BACKGROUND YLINDRICAL and curved roof buildings are increasingly adopted in modern agricultural architecture as they offer aerodynamically efficient shapes and provide designers with an alternative to rectangular building forms. Nevertheless, as observed by Blackmore et al. [1], there is little information available on the wind loads on cylindrical roofs. The proposed Eurocode for wind actions [2] includes pressure coefficients for a limited range of aspect ratio cylindrical roofs, obtained from experimental measurements in low-turbulence conditions, but only for wind blowing normal to the eaves. Some other national wind codes, such as the Australian and New Zealand code [3], the Canadian one [4] and the American (ASCE) one [5] provide only external pressure coefficients for curved roofs. Some other data are also reported by Cook [6] and Blackmore et al. [1]. The limitations in accurate pressure coefficient databases can be overcome by using advanced CFD (Computational Fluid Dynamic) codes, which can outflank the lack of experimental data thanks to their inherent ability to determine the aerodynamic components of actions through the Marco Raciti Castelli is Research Associate at the Department of Mechanical Engineering of the University of Padua, Via Venezia 1, 35131 Padova, Italy and fluid dynamic specialist at ESPE S.r.l., Via Cappello 12/A, 35010 San Pietro in Gu (PD) (e-mail: marco.raciticastelli@unipd.it). Sergio Toniato is Executive and Design Manager at ESPE S.r.l., Via Cappello 12/A, 35010 San Pietro in Gu (PD), Italy (e-mail: stoniato@espe.it). integration of the Navier-Stokes equations in the neighborhood of the building. As observed by Raciti Castelli et al [7], the use of commercial CFD packages to calculate the wind flow and resulting action on civil structures has aroused a large credit both in research and academic communities as well as in consulting engineering societies, thanks to their capability of providing an insight into the flow field around the buildings even before their construction. Nevertheless, as observed by Blackmore et al. [1], the modeling of atmospheric turbulence, interacting with the structure-generated turbulent flow still depicts large difficulties that often lead to erroneous results when areas of flow separation are to be simulated. These problems have been discussed by several authors, including in particular Stathopoulos [8] [9], Timofeyef [10] and Ferreira et al. [11]. As pointed out by Stathopoulos [12], the flow around buildings is still extremely difficult to predict by computational methods, even for simple surrounding environments. However, there is increasing evidence that CFD-based techniques provide adequate responses in case the mean flow and pressure conditions are to be determined. Some of the specific design issues in the use of CFD for the prediction of flows around buildings were described by several authors: • Yoshie et al. [14] performed comparative and parametric studies on the flow around a square prism, based on the work of Meng and Hibi [15], in order to validate CFD simulations of the absolute velocity field around high-rise buildings; • Hak-Sun et al. [15] presented a numerical simulation of turbulent wind flow around a complex building using LES (Large Eddy Simulation). The numerical results were validated against the experimental measurements of a multi-block configuration of the WERFL (Wind Engineering Research Field Laboratory) building at Texas Tech University [16], showing good agreement; • Raciti Castelli et al. [17] investigated the flow field over a flat roof model building, in order to assess CFD guidelines for the calculation of the turbulent flow over a structure immersed in an atmospheric boundary layer. To this purpose, a complete validation campaign was performed through a systematic comparison of numerical simulations and wind tunnel experimental data. Several Ernesto Benini is Associate Professor at the Department of Mechanical Engineering of the University of Padua, Via Venezia 1, 35131 Padova, Italy (e-mail: ernesto.benini@unipd.it). Marco Raciti Castelli, Sergio Toniato and Ernesto Benini Numerical Analysis of Wind Loads on a Hemicylindrical Roof Building C World Academy of Science, Engineering and Technology International Journal of Mechanical and Mechatronics Engineering Vol:5, No:8, 2011 1669 International Scholarly and Scientific Research & Innovation 5(8) 2011 scholar.waset.org/1307-6892/8529 International Science Index, Mechanical and Mechatronics Engineering Vol:5, No:8, 2011 waset.org/Publication/8529