بهينه سازي تركيبي ايرفويل در جريان رينولدز فوق پايين برمبناي طراحي معكوس، مدل جايگزين و الگوريتم ژنتيك در زاويه حمله بهينه به همراه ابداع ايرفويل حفره دار هندسه متغير در زواياي حمله بالا

This study investigates the shape optimization an‎d aerodynamic performance enhancement of airfoils operating in an ultra-low Reynolds number flow regime (Re = 1000). The research is divided into two main parts: (1) optimization of the airfoil geometry at a specific angle of attack of 7 degree, an‎d (2) enhancement of airfoil performance over an angle-of-attack range of 12 to 26 degree to improve maneuverability. In the first part, airfoil shape optimization at an angle of attack of 7 degree was performed using an inverse-design approach based on a shear-stress-driven elastic-shell algorithm to generate the initial dataset, followed by a genetic algorithm to conduct the optimization process. In the second part, streamline-based geometric modifications were applied over the angle-of-attack range of 12 to 26 degree to control the unsteady separated flow over the airfoil. In the first stage of the study, an extensive dataset containing a wide range of airfoil geometries an‎d their corresponding aerodynamic coefficients was generated using an enhanced shear-stress-based elastic-shell inverse-design algorithm. The enhancement involved allowing the vertical displacement of the beam’s leading an‎d trailing nodes, thereby enabling the generation of diverse an‎d arbitrary geometries. Shear stress was selec‎ted as the governing parameter for inverse design because of the low flow inertia an‎d the relatively low sensitivity of the pressure parameter in this flow regime. Following dataset generation, a surrogate model based on an interaction regression formulation was trained to establish a direct mapping between the geometric parameters an‎d the aerodynamic coefficients. Predicting aerodynamic coefficients directly from geometric characteristics provides high sensitivity to geometric variations an‎d enables their effects to be directly reflected in the aerodynamic quantities. Furthermore, this approach facilitates the analysis an‎d differentiation of a large an‎d diverse population of geometries. Therefore, the direct mapping of airfoil geometry to aerodynamic coefficients represents a suitable approach for optimization an‎d exploration in insufficiently understood flow regimes. Subsequently, the generated dataset an‎d the surrogate model were incorporated into a genetic optimization algorithm. Two objective functions were considered: maximizing the lift-to-drag ratio an‎d minimizing the drag coefficient at a prescribed constant lift coefficient. A thickness constraint was also imposed to prevent excessive thinning of the optimized geometries. The results obtained using the lift-to-drag-ratio maximization objective an‎d the constant-lift drag-minimization objective demonstrated a 24.83% increase in the lift-to-drag ratio an‎d a 6.27% reduction in the drag coefficient, respectively. In the second part, the aerodynamic performance of three airfoils—namely, the NACA 4404 airfoil an‎d the two optimized geometries obtained from the first part using the aforementioned objective functions—was improved over the angle-of-attack range of 12 to 26 degree by introducing streamline-based geometric modifications on the suction surface.

مهدي نيلي احمدآبادي

محسن ثقفيان , علي انصاري