Three dimensional design optimization using analytical and numerical jacobians

Date

2011-12-01

Author

Eyi, Sinan
Degirmenci, M.
Yumusak, M.

Metadata

Show full item record

Item Usage Stats

162
views

0
downloads

A design optimization method based on three dimensional Euler equations is developed. A finite volume method is implemented to discretize the Euler equations. Newton's method is used to solve the discretized form of Euler equations. Newton's method requires the calculation of the Jacobian matrix which is the derivative of the residual vector with respect to the flux vector. Different upwind methods are used in the calculation of flux vectors. Numerical and analytical methods are utilized in the evaluation of Jacobian matrices. The efficiency and accuracy of the analytical and numerical Jacobian evaluations are compared. In order to improve the accuracy of numerical method, detailed error analyses are performed. The optimum finite difference perturbation magnitude that minimizes the error is searched. The computation time of numerical Jacobian evaluation is reduced by calculating the flux vectors with perturbed flow variables only in related cells. The performances of different sparse matrix solvers are also compared. The effects of errors in numerical Jacobians on the accuracy of sensitivities is analyzed. Results show that the finitedifference perturbation magnitude and computer precision are the most important parameters that affect the accuracy of numerical Jacobians. Approximately the same optimum perturbation magnitude enables the most accurate numerical flux Jacobian and sensitivity calculations. © 2011 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

URI

https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84880629151&origin=inward
https://hdl.handle.net/11511/80282

Conference Name

20th AIAA Computational Fluid Dynamics Conference (2011)

Collections

Department of Aerospace Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Three dimensional rocket nozzle design using adjoint method Eyi, Sinan (2013-09-16) A design optimization method based on three dimensional Euler equations is developed. A finite volume method is implemented to discretize the Euler equations. Newton's method is used to solve the discretized form of Euler equations. Newton's method requires the calculation of the Jacobian matrix which is the derivative of the residual vector with respect to the flux vector. Different upwind methods are used in the calculation of flux vectors. Numerical and analytical methods are utilized in the evaluation o...
Three dimensional rocket nozzle design using adjoint method Eyi, Sinan (2013-01-01) A design optimization method based on three dimensional Euler equations is developed. A finite volume method is implemented to discretize the Euler equations. Newton’s method is used to solve the discretized form of Euler equations. Newton’s method requires the calculation of the Jacobian matrix which is the derivative of the residual vector with respect to the flux vector. Different upwind methods are used in the calculation of flux vectors. Numerical and analytical methods are utilized in the evaluation o...
Aerodynamic design optimization of three dimensional rocket nozzles using adjoint method Eyi, Sinan (null; 2013-09-13) A design optimization method based on three dimensional Euler equations is developed. A finite volume method is implemented to discretize the Euler equations. Newton's method is used to solve the discretized form of Euler equations. Newton's method requires the calculation of the Jacobian matrix which is the derivative of the residual vector with respect to the flux vector. Different upwind methods are used in the calculation of flux vectors. Numerical and analytical methods are utilized in the evaluation o...
Two dimensional finite volume weighted essentially non-oscillatory euler schemes with different flux algorithms Aktürk, Ali; Akmandor, İbrahim Sinan; Department of Aerospace Engineering (2005) The purpose of this thesis is to implement Finite Volume Weighted Essentially Non-Oscillatory (FV-WENO) scheme to solution of one and two-dimensional discretised Euler equations with different flux algorithms. The effects of the different fluxes on the solution have been tested and discussed. Beside, the effect of the grid on these fluxes has been investigated. Weighted Essentially Non-Oscillatory (WENO) schemes are high order accurate schemes designed for problems with piecewise smooth solutions that invol...
Two dimensional finite volume weighted essentially non-oscillatory euler schemes with uniform and non-uniform grid coefficients Elfarra, Monier Ali; Akmandor, İbrahim Sinan; Department of Aerospace Engineering (2005) In this thesis, Finite Volume Weighted Essentially Non-Oscillatory (FV-WENO) codes for one and two-dimensional discretised Euler equations are developed. The construction and application of the FV-WENO scheme and codes will be described. Also the effects of the grid coefficients as well as the effect of the Gaussian Quadrature on the solution have been tested and discussed. WENO schemes are high order accurate schemes designed for problems with piecewise smooth solutions containing discontinuities. The key ...

Citation Formats

S. Eyi, M. Degirmenci, and M. Yumusak, “Three dimensional design optimization using analytical and numerical jacobians,” Honolulu, HI; United States, 2011, Accessed: 00, 2021. [Online]. Available: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84880629151&origin=inward.