Large eddy simulation of wind loading on an anticlastic conical tensile membrane

• Published in: Journal of wind engineering and industrial aerodynamics 2024-03, Vol.246, p.105658, Article 105658
• Main Authors: De, Budhaditya, Kumar, Ajay, Mishra, Sudib Kumar
• Format: Article
• Language: English
• Subjects: Aerodynamic roughness; Coherence function; Computational fluid dynamics (CFD); Large eddy simulation; Power spectral density; Tensile membrane structures; Wind tunnel testing
• ISSN: 0167-6105; 1872-8197
• DOI: 10.1016/j.jweia.2024.105658

Wind loading on Tensile Membrane Structure(s) (TMSs) is conventionally investigated by the Wind Tunnel Testing (WTT) and is complemented by the Reynold's-Averaged-Navier-Stokes (RANS) based simulation, which lacks fine flow features. To capture finer flow features, this study employs the Large Eddy Simulation (LES) to investigate an anticlastic conical TMS, with and without façade. The sub-grid viscosity is modelled by the Wall Adapting Local Eddy (WALE) model. Ninety per cent of the Turbulent Kinetic energy (TKE) is shown to be resolved (via Pope Criteria) by the LES. The pertinent flow fields and the influence of parametric variations (aerodynamic roughness, rise-to-span ratio of the TMS and reference wind speed) are presented using the streamlines, velocity, vorticity and pressure maps. The pressure coefficients on the TMS surface are summarized based on zonation of the membrane surface. The pressure time histories are analyzed for their Gaussian/non-Gaussianity. The power spectral density (PSD) model is presented for the fluctuating wind pressure coefficients in an existing three-parameter model and a decaying exponential coherence function, accounting for the inter-zonal cross-spectral density, which may be useful for simulation and stochastic analysis. The zone-wise Root Mean Square (RMS) values and the peak factors are tabulated for use in design. •Large Eddy Simulation of aerodynamics around conical tensile membrane is presented.•Pertinent flow fields are presented using streamlines, vorticity and pressure map.•Spatio-temporal pressure distribution is characterized on the membrane surface.•A clustering based zonation is presented for the design and peak pressure coefficient.•Stochastic model is presented for describing the multi-point wind excitations.