URN etd-0614105-095254 Statistics This thesis had been viewed 3002 times. Download 466 times. Author Mei-Hsia Chang Author's Email Address No Public. Department Mechanical Engineering Year 2004 Semester 2 Degree Ph.D. Type of Document Doctoral Dissertation Language English Page Count 223 Title Development and Applications of Inverse Heat Transfer Method Keyword Temperature prediction Simplified conjugate-gradient method Shape identification Shape design Inverse heat transfer Conjugate-gradient method Automatic-filter scheme Automatic-filter scheme Conjugate-gradient method Inverse heat transfer Shape design Shape identification Simplified conjugate-gradient method Temperature prediction Abstract The aim of this thesis is to develop inverse heat transfer methods and apply these methods to optimal design problems in different areas. The inverse heat transfer methods developed in the present thesis include the conjugate-gradient method (CGM), simplified conjugate-gradient method (SCGM), and automatic-filter scheme (AFS). Firstly, the CGM method is modified and categorized into two types: polynomial-function approach and point-by-point approach to extend its feasibility in (1) shape identification for the ice layer within the cylindrical capsules in cold storage system based on the temperature data on the outer surface of these capsules, (2) predictions of inner temperature distribution at the interface between the carbon plate and the membrane electrode assembly of a proton exchange membrane fuel cell (PEMFC), and (3) shape design of slider surfaces to meet the specified load demands.
The CGM is limited in the definition of objective function form. Thus it is not applied to problems with a more general objective function definition. In order to overcome the limitation, the SCGM method is presented by modifying the CGM method. The capability of the SCGM in identifing the boundary of the inner voids in the solid bodies based on the thermal data is tested by several test cases.
In addition, in the thesis a novel inverse heat transfer method is proposed, which incorporates an automatic-filter scheme (AFS) with the CGM method, for identifying shapes, positions, and temperature of heating elements embedded in a rectangular package. The identification is simply based on the upper surface temperature data of the rectangular package. Shapes of the heating elements are visualized by using node-matrix images. In this manner, numerical grid generation is not required and without overwhelming mathematical manipulation, and therefore the form of objective function with this method becomes more flexible.Advisor Committee Chin-Hsiang Cheng - advisor
Chih-yung Wen - co-chair
Ching-Jenq Ho - co-chair
Jeng-Rong Ho - co-chair
Pei-Yuan Tzeng - co-chair
Wei-Mon Yan - co-chair
Yau-Ming Chen - co-chair
Files indicate access worldwide Date of Defense 2005-06-02 Date of Submission 2005-06-14