Multifidelity aerodynamic shape optimization for mitigating dynamic stall using Cokriging regression-based infill

This work proposes a multifidelity modeling approach to mitigate adverse characteristics of airfoil dynamic stall through aerodynamic shape optimization (ASO). Cokriging regression (CKR) is used to efficiently determine an optimum airfoil shape by combining data from high-fidelity (HF) and low-fidelity (LF) computational fluid dynamics simulations. The HF dynamic stall response is modeled using the unsteady Reynolds-averaged Navier–Stokes equations and Menter’s SST turbulence model, whereas the LF model is developed by simplifying the HF model with a coarser discretization and relaxed convergence criteria. The CKR model, constructed using various infill criteria to model the objective and constraint functions with six PARSEC parameters, is utilized to find the optimal design. The results show that the optimal shape from CKR delays the dynamic stall angle over 3° while reducing the peak values of the aerodynamic coefficients compared to the baseline airfoil (NACA 0012). Comparing the optimized shapes from the CKR and a HF Kriging regression (HF-KR) shows a similar delay in dynamic stall angle; however, the CKR optimum provides a better design for the current problem formulation while requiring 39% less computational time than the HF-KR approach. This work presents a new multifidelity modeling approach to saving the computational burden of dynamic stall mitigation through ASO. The approach used in this work is general and can be applied for other unsteady aerodynamic applications and optimization.