Parametric modeling and automated aerodynamic evaluation of large civil aircraft nose based on NURBS
Abstract
To address the challenges of complex geometric constraints, large parametric sample sizes, and high computational cost of full-aircraft CFD in early-stage nose aerodynamic design of large passenger aircraft, this paper constructs an equivalent isolated-nose CFD automation framework based on NURBS parameterization. First, with the nose tip as the coordinate origin, the nose geometry is described by characteristic curves including the upper zero-longitudinal curve, the lower zero-longitudinal curve, the maximum-width curve, and transverse profile curves, with 9 macroscopic parameters and 29 normalized parameters controlling the geometric variations. Subsequently, parametric geometry, computational domain reconstruction, boundary naming, mesh generation, and steady flow-field solving are integrated into a headless Python-based automated workflow. Numerical computations employ the starboard half-model, pressure far-field boundaries, no-slip adiabatic wall conditions, the Ideal-Gas model, and the SST k-omega turbulence model. Grid sensitivity and aft-extension length sensitivity analyses are conducted at the Ma = 0.85 cruise condition. Results show that the total local nose drag across the four grid levels ranges from 226,728.22 N to 229,300.76 N, with an overall variation of approximately 1.1346%; the 3.57-million-cell mesh differs from the adjacent coarser and medium-fine meshes by 0.3117% and 0.3188%, respectively, and from the 21.92-million-cell fine mesh by 0.8204%. The aft-extension length sensitivity results indicate that the total local drag variation from the 3L to the 5L extension is only 0.0848%. Under the current CFD setup and the total local drag metric, the 3.57-million-cell mesh and the 3L aft-extension can serve as a viable configuration for batch automated parametric evaluation.