Adjoint-based Shape Sensitivities for Turbomachinery Design Optimizations
نویسندگان
چکیده
An adjoint preprocess for an adjoint-based turbomachinery design process is described. The resulting process is compared to a previously established adjoint process based on three-dimensional adjoint solutions and deformed meshes. Within the new process, the calculation of sensitivities is performed using shape sensitivities and surface displacements, where the design parameters are computer-aided design (CAD) parameters describing the geometry of a turbomachinery component. In particular, an automatized process for the generation of surface displacements from variations of the CAD parameters is described. Shape sensitivities are calculated using an adjoint elliptic mesh deformation tool. The method is applied to a counterrotating fan. Integrating the adjoint preprocess into the gradient evaluation results in a significantly reduced memory usage, since no perturbed three-dimensional meshes have to be generated. It also has the potential to reduce the dependency of the computation time on the number of design parameters. Moreover, the shape sensitivities provide interesting insights into the design problem.
منابع مشابه
C:/Documents and Settings/ssharma/Desktop/ECMI07/ECMI07/gauger.dvi
Because detailed aerodynamic shape optimizations still suffer from high computational costs, efficient optimization strategies are required. Regarding the deterministic optimization methods, the adjoint approach is seen as a promising alternative to the classical finite difference approach [1, 2]. With the adjoint approach the sensitivities needed for the aerodynamic shape optimization can be e...
متن کاملAdjoint Algorithm for CAD-Based Shape Optimization Using a Cartesian Method
We focus on the computation of objective function gradients using a discrete adjoint method for embedded-boundary Cartesian meshes. This is a significant step in our research toward using computer-aided design (CAD) directly for adjoint-based gradient computations. The approach treats the Cartesian mesh as a rigid structure. Under this assumption, accurate evaluation of mesh sensitivities depen...
متن کاملAdjoint Algorithm for CAD-Based Shape Optimization Usign a Cartesian Method
We focus on the computation of objective function gradients using a discrete adjoint method for embedded-boundary Cartesian meshes. This is a significant step in our research toward using computer-aided design (CAD) directly for adjoint-based gradient computations. The approach treats the Cartesian mesh as a rigid structure. Under this assumption, accurate evaluation of mesh sensitivities depen...
متن کاملMultidisciplinary Optimization in Turbomachinery Design
This paper investigates particular aspects of performing multidisciplinary optimizations in different turbomachinery design steps. The differences in the optimization approaches and used methods in preliminary and final design steps are shown. An optimization environment is developed, which supports multidisciplinary turbomachinery design. The general concept and components are explained. Some ...
متن کاملNon-parametric Shape Optimization in Industrial Context
Non-parametric shape optimization has its roots in the work from Zienkienwicz et al. [17] in 1973. In 1979 Schnack [15] published a sensitivity-free optimality criteria method for shape stress optimization which was the basis for TOSCA Structure.shape [10, 14] which even today solves difficult industrial non-parametric shape optimization problems. The TOSCA Structure technology is now also avai...
متن کامل