A new method for gradient-based flutter constrained optimization

A new method for gradient-based flutter constrained optimization is presented. An evaluation of the aeroelastic phenomenon of flutter through the control of the damping is studied. Through the mathematical formulation of the flutter is created a flutter constraint to pose the basis for a gradient-based structural optimization with as objective the mass objective, and as constraints stress and flutter. The method is intended to be used in a gradient-based analysis which allows solving large optimization in reasonable computational time. Structural optimization methods must provide a solution in a limited time for making effective use of high-performance computing resources. Besides, in adjoint-based gradient methods is necessary to obtain the gradient, with respect to thousands of design variables and this action has a cost similar to the analysis. In this framework, it is used a gradient-based optimization method and it is formulated a way for taking the derivatives of formulation of the flutter. In the research are used programs of structural optimization that have the advantage of being open-source, free, and under continuum development. Furthermore, the Python language on which are based the programs is a well-defined programming language and is very flexible. These aspects allow the possibility of further improvements. The arising of the phenomenon of flutter on aircraft’s wing is controlled through the eigenvalues of the aeroelastic system. A flutter constraint is formulated including a constraint on the value of the real part of the eigenvalues for several different points in the flight envelope. The objective is to create an aeroelastic constraint and take the derivative of its mathematical formulation, having a function that can predict and control the flutter and possibility of the switching of the modal shapes, the phenomenon that causes a switch of the point of flutter to a different modal shape and a different velocity.