Accuracy of the finite difference sensitivity calculations are improved by
calculating the optimum finite difference interval sizes. In an aerodynamic inverse
design algorithm, a compressor cascade geometry is perturbed by shape functions
and finite differences sensitivity derivatives of the flow variables are calculated with
respect to the base geometry flow variables. Sensitivity derivatives are used in an
optimization code and a new airfoil is designed verifying given design
characteristics. Accurate sensitivities are needed for optimization process. In order to
find the optimum finite difference interval size, a method is investigated.
Convergence error estimation techniques in iterative solutions and second derivative
estimations are investigated to facilitate this method. For validation of the method,
analytical sensitivity calculations of Euler equations are used and several
applications are performed.
Efficiency of the finite difference sensitivity calculations is improved by
parallel computing. Finite difference sensitivity calculations are independent tasks in
an inverse aerodynamic design algorithm and can be computed separately.
Sensitivity calculations are performed on parallel processors and computing time is
decreased.
| Identifer | oai:union.ndltd.org:METU/oai:etd.lib.metu.edu.tr:http://etd.lib.metu.edu.tr/upload/12609128/index.pdf |
| Date | 01 December 2007 |
| Creators | Ozhamam, Murat |
| Contributors | Eyi, Sinan |
| Publisher | METU |
| Source Sets | Middle East Technical Univ. |
| Language | English |
| Detected Language | English |
| Type | M.S. Thesis |
| Format | text/pdf |
| Rights | To liberate the content for public access |
Page generated in 0.0081 seconds