Surrogate-based aeroelastic loads prediction in the presence of shock-induced separation

Abstract A surrogate-based modeling strategy is presented for robust and efficient prediction of unsteady aeroelastic loads in the presence of shock-induced separation. Enriched piston theory predictions are extended with a data-driven nonlinear autoregressive with exogenous inputs model to account for hysteresis from the interplay of a dynamically deforming surface with the separation bubble in a shock/boundary layer interaction. The approach is evaluated for prescribed surface motion and shock-induced panel flutter responses, with good agreement observed in each scenario relative to unsteady Reynolds-averaged Navier–Stokes simulations. For the latter, excellent agreement is observed in the prediction of the stability boundary and oscillation frequency. In contrast, the oscillation amplitude conservatively deviates from the Reynolds-averaged Navier–Stokes solution with increasing dynamic pressure. The online computational cost of the extended approach is orders of magnitude less than that required for predictions using an unsteady Reynolds-averaged Navier–Stokes model.