Multidisciplinary Automated Optimization Strategy on a Counter Rotating Fan

Abstract

Multidisciplinary automated optimization processes are nowadays essential to obtain optimal turbomachinery components. However, an extremely large number of free variables, constraints and objectives results in a complex task. This paper presents an optimization strategy developed to handle with different constraints, design goals concerning aerodynamic, mechanic and aeroelastic, and finally manufacturing aspects. This strategy has been applied to a counter rotating integrated shrouded propfan, which is developed within a DLR-project. Both rotors have been already aerodynamic optimized in a first design phase coupled with a mechanical analysis of the CF/PEEK blades with titanium clevises. Detailed analysis showed high displacements and unreliable Campbell-Diagrams. To reach a rig-ready design a new optimization strategy has been developed. The optimizations feature more than hundred free variables, two objective functions, as well as a high number of aerodynamic and mechanical constraints. The mechanical behavior of the blades has been improved step by step in four successive aeromechanical optimizations. To secure the improvement obtained in one optimization, their objective functions become constraints in the next step. In the first optimization, the efforts have been focused on reducing the maximal absolute displacements in several operation points. In the second one, the scattering of the maximal absolute displacements between several operation points have been reduced. In the third optimization, the Campbell-Diagrams have been additionally optimized. Although the aerodynamic performance remained on a good level, it decreased a little bit in this design phase. For this reason, an additional fourth optimization was performed with the objective to increase the fan efficiency by keeping the good mechanical behavior reached before. The presented optimization strategy has been successfully completed and the best members obtained show an almost satisfactory mechanical feasibility in view of the planned rig test.