Loading...

Three-dimensional Aerothermal Shape Optimization of Turbine Blade, Using Gradient Adjoint-based Methods

Zeinalpour, Mehdi | 2018

1255 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 50768 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Mazaheri, Karim
  7. Abstract:
  8. The most important challenge and the most time consuming part of the gradient based optimization algorithms in the aerodynamic shape optimization problems is the evaluation of the sensitivity of the objective function with respect to the design variables. The adjoint method which has been the subject of many research in the recent three decades, is capable of computing the complete gradient information needed for optimization by solving the governing flow equations and their corresponding adjoint equations only once, regardless of the number of design parameters. In this study, the continues adjoint equations for compressible inviscid and viscous flows are derived and the formulation of adjoint boundary condition for the minimization of entropy generation in internal flow problems such as turbine blade is presented in details. The new adjoint system which is written in the global Cartesian coordinate system can be applied to problems with both structured and unstructured grids. Also, in this thesis the adjoint method is extended to the conjugate heat transfer problems in which the compressible viscous flow field is coupled to heat transfer in the solid region. The new coupled adjoint formulation can be used in problems such as aerodynamic optimization of the cooled turbine blade, turbine blade cooling systems and heat exchangers in which the working fluid flow is compressible. The gradients computed using the presented adjoint systems are compared with those obtained with finite difference method in different optimization problems such as: turbine blade inversed design with both inviscid and viscous flows, entropy minimization in turbine blade with both inviscid and viscous flows and inversed design of internal cooling channel. The results show a good agreement between the gradients computed using adjoint and finite difference methods. The capability of the proposed formulation is shown by solving some two and three dimensional optimization problems
  9. Keywords:
  10. Conjugate Heat Transfer ; Adjoint Method ; Shape Optimization ; Heat Transfer ; Turbine Blades ; Axial Turbine Blade

 Digital Object List

 Bookmark

...see more