Scholarlink

<p>Morphing wings possess the unique capability to alter their aerodynamic characteristics through shape modifications during flight, enabling enhanced maneuverability and adaptability to varying flight dynamics. This study introduces an innovative aerodynamic design methodology for the RAE2822 subsonic wing (0.3 ≤ Mach ≤ 0.5) featuring trailing-edge morphing, tailored for diverse flight scenarios. Central to this approach is the development of a morphing function that establishes a quantitative relationship between geometric parameters and aerodynamic characteristics across multiple flight conditions. This function operates as a learning system, allowing for the prediction of aerodynamic outputs based on independent input variables, including geometric parameters and flight conditions. The morphing function is derived through surrogate modeling of the response surface, trained using 2265 numerical design of experiment tests that incorporate three geometric and two flight variables, facilitating a comprehensive sensitivity analysis. Morphing shapes are generated via a polynomial perturbation function designed to satisfy specific flow conditions. Aerodynamic specifications, including lift coefficient (CL) and drag coefficient (CD), are computed for each configuration using a Reynolds-Averaged Navier–Stokes (RANS) computational fluid dynamics approach, which integrates continuous geometry manipulations and a dynamic mesh morpher. Following the creation of 151 morphing shapes and the evaluation of their aerodynamic characteristics across 15 flight conditions, a multi-dimensional response surface was established. The surrogate models for aerodynamic specifications were analyzed using three methodologies: polynomial singular value decomposition (SVD), Kriging, and radial basis function (RBF), with 80% of the data utilized for model training. Results demonstrated that RBF and Kriging models exhibited superior predictive accuracy for aerodynamic coefficients, achieving an R-squared value of 0.998. Subsequently, a multi-scenario trajectory optimization plan was developed, employing morphing functions as surrogate meta-models within a Multi-Objective Genetic Algorithm framework. This optimization successfully identified optimal morphing geometries and flight condition variables, facilitating enhanced performance across multiple CL and CD trajectory scenarios. Additionally, CFD validation of a sample take-off command shape predicted by the morphing function reported a 3.8% error in the CL value which reinforced morphing function idea in the morphing wing design process.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Aerodynamic Design of a Trailing-Edge Morphing Wing Using Morphing Function Idea and Meta-model Optimization Based on CFD Approach

  • 期刊论文

  • 发表日期：

    2026年1月7日

查看全文
  • Hadi Ziatabar,
  • Mohammad Taeibi Rahni,
  • Arash Shams Taleghani

摘要

Morphing wings possess the unique capability to alter their aerodynamic characteristics through shape modifications during flight, enabling enhanced maneuverability and adaptability to varying flight dynamics. This study introduces an innovative aerodynamic design methodology for the RAE2822 subsonic wing (0.3 ≤ Mach ≤ 0.5) featuring trailing-edge morphing, tailored for diverse flight scenarios. Central to this approach is the development of a morphing function that establishes a quantitative relationship between geometric parameters and aerodynamic characteristics across multiple flight conditions. This function operates as a learning system, allowing for the prediction of aerodynamic outputs based on independent input variables, including geometric parameters and flight conditions. The morphing function is derived through surrogate modeling of the response surface, trained using 2265 numerical design of experiment tests that incorporate three geometric and two flight variables, facilitating a comprehensive sensitivity analysis. Morphing shapes are generated via a polynomial perturbation function designed to satisfy specific flow conditions. Aerodynamic specifications, including lift coefficient (CL) and drag coefficient (CD), are computed for each configuration using a Reynolds-Averaged Navier–Stokes (RANS) computational fluid dynamics approach, which integrates continuous geometry manipulations and a dynamic mesh morpher. Following the creation of 151 morphing shapes and the evaluation of their aerodynamic characteristics across 15 flight conditions, a multi-dimensional response surface was established. The surrogate models for aerodynamic specifications were analyzed using three methodologies: polynomial singular value decomposition (SVD), Kriging, and radial basis function (RBF), with 80% of the data utilized for model training. Results demonstrated that RBF and Kriging models exhibited superior predictive accuracy for aerodynamic coefficients, achieving an R-squared value of 0.998. Subsequently, a multi-scenario trajectory optimization plan was developed, employing morphing functions as surrogate meta-models within a Multi-Objective Genetic Algorithm framework. This optimization successfully identified optimal morphing geometries and flight condition variables, facilitating enhanced performance across multiple CL and CD trajectory scenarios. Additionally, CFD validation of a sample take-off command shape predicted by the morphing function reported a 3.8% error in the CL value which reinforced morphing function idea in the morphing wing design process.