A parabolic velocity-decomposition method for wind turbines
详细信息 查看全文
- 作者：Anshul Mittal ; Anshul-Mittal@mocs.utc.edu ; W. Roger Briley ; Kidambi Sreenivas ; Lafayette K. Taylor
- 关键词：PVD ; PNS ; Parabolic ; Parabolized ; Wind turbine wake ; Navier&ndash ; Stokes
- 刊名：Journal of Computational Physics
- 出版年：2017
- 出版时间：1 February 2017
- 年：2017
- 卷：330
- 期：Complete
- 页码：650-667
- 全文大小：2656 K
- 卷排序：330
文摘
An economical parabolized Navier–Stokes approximation for steady incompressible flow is combined with a compatible wind turbine model to simulate wind turbine flows, both upstream of the turbine and in downstream wake regions. The inviscid parabolizing approximation is based on a Helmholtz decomposition of the secondary velocity vector and physical order-of-magnitude estimates, rather than an axial pressure gradient approximation. The wind turbine is modeled by distributed source-term forces incorporating time-averaged aerodynamic forces generated by a blade-element momentum turbine model. A solution algorithm is given whose dependent variables are streamwise velocity, streamwise vorticity, and pressure, with secondary velocity determined by two-dimensional scalar and vector potentials. In addition to laminar and turbulent boundary-layer test cases, solutions for a streamwise vortex-convection test problem are assessed by mesh refinement and comparison with Navier–Stokes solutions using the same grid. Computed results for a single turbine and a three-turbine array are presented using the NREL offshore 5-MW baseline wind turbine. These are also compared with an unsteady Reynolds-averaged Navier–Stokes solution computed with full rotor resolution. On balance, the agreement in turbine wake predictions for these test cases is very encouraging given the substantial differences in physical modeling fidelity and computer resources required.