Loading...
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 41774 (08)
- University: Sharif University of Technology
- Department: Mechanical Engineering
- Advisor(s): Farhanieh, Bijan
- Abstract:
- Tomography offers a unique opportunity to unravel the complexities of the internal structure of an object without the need to invade it. There are various methods for this reason, in which all of them occupy wave or energy,ffiduction for imaging inside an object. Since the tomography plays a vital role in medicine and industry and present methods have limitation and difficulties to operate, the necessity for developing it has been inevitable. In this thesis a new tomography method has been presented in which uses inverse heat conduction for estimating unknown object with heat conductivity k2 inside a known domain with heat conductivity k 1 • For this purpose temperature is measured at some sensor points on external boundary with specific boundary condition, then a shape is guessed and the• related temperatures at those sensor points are calculated, the Levenberg-Marquardt algorithm and the conjugate gradient method is used for minimizing the error between measured and estimated temperatures. Geometry of internal object is approximated with an eight-point third order closed spline. For simulating heat conduction with CFD, boundary fitted circular mesh is used. The resulting mesh is a non orthogonal mesh, so the term of on-orthogonality of grids has been added to governing equation Various case studies have been investigated to ensure the reliability of this method. Results have been shown •that Levenberg-Marquardt method could find any smoothed object in a few iterations.
- Keywords:
- Inverse Heat Transfer ; Image Reconstruction ; Computational Fluid Dynamics (CFD) ; Spline Method ; Tomography ; Control Volume ; Inverse Problem ; Nonorthogonal Boundary Fitted
-
محتواي پايان نامه
- view