Dataset for the paper "Parametric Nonlinear Volterra Series via Machine Learning: Transonic Aerodynamics"

This dataset provides high-fidelity Computational Fluid Dynamics (CFD) simulations used in:G. Immordino, A. Da Ronch, and M. Righi. "Parametric Nonlinear Volterra Series via Machine Learning: Transonic Aerodynamics." Journal of Aircraft, 2025.https://arc.aiaa.org/doi/abs/10.2514/1.C038288 The datasets support the identification and validation of linear and nonlinear Volterra series kernels for unsteady transonic aerodynamics. NACA0012 Airfoil (2D case): Generated using SU2 v7.5.0 with the Unsteady RANS (URANS) formulation and Spalart–Allmaras turbulence model (negative production option active). Structured O–mesh with ~177k elements; y+≈1; domain extended to 100 chords. Simulations employed Δτ=0.15 over a total nondimensional time of 113.6. Step responses were induced by plunge motions of −1 m/s and −2 m/s, covering Mach numbers M=[0.58 , 0.76] and angles of attack AoA=[0 , 7] deg. Output includes time-resolved lift and pitching moment coefficients (CL , CM). Benchmark Supercritical Wing – BSCW (3D case): Semi–span wing configuration from the AIAA Aeroelastic Prediction Workshop, simulated using SU2 v7.5.0 URANS. Mixed grid with 15.6M elements (structured near wing and boundary layer, voxelised in the farfield); y+≈1. Nondimensional timestep 0.029; total nondimensional time =27.2. Pitch step responses with amplitudes of 1 deg and 2 deg, spanning Mach numbers M=[0.70 , 0.84] and angles of attack AoA=[0 , 5] deg. Dataset includes aerodynamic loads (CL , CM), with flow regimes covering attached flow, onset of shock-induced separation, and dynamic stall. PurposeThe data enable reduced-order modelling of unsteady transonic aerodynamics through Volterra series expansions, and support studies in aeroelasticity, flutter prediction, and machine-learning–based surrogate modelling. Contents Step–response histories for multiple Mach–AoA combinations. Linear and nonlinear excitation cases (small vs. large input amplitudes). Keywords:Transonic aerodynamics, unsteady aerodynamics, Volterra series, reduced-order models, flutter, NACA0012, Benchmark Supercritical Wing, CFD, machine learning.