| Summary: | 碩士 === 國立成功大學 === 造船及船舶機械工程學系 === 87 === An inverse design problem is solved to determine the shape of complex coolant flow passages in internal cooled turbine blades by using the conjugate gradient method (CGM). One of the advantages of using CGM lies in that it easily can handle problems having a huge number of unknown parameters and converges very fast. The boundary element method (BEM) is used to calculate the direct, sensitivity and adjoint problems due to its characteristics of easily-handling the problem considered here.
Results obtained by using the CGM to solve the inverse problems are verified based on the numerical experiments in the analysis model. One concludes that the CGM is applied successfully in estimating the arbitrary shape of cavities and the rate of convergence is also very fast even when the number of unknown parameters is large. Moreover, the design model of the inverse problem is also performed to estimate the optimal shape of cooling passages in accordance with the desired blade surface temperature distributions..
The applications of inverse problems can be found in several engineering fields, such as the determination of boundary conditions, thermal properties, heat generation, contact resistance and thermal properties and thickness of frost on the tube of evaporator [1-4]. The purpose of the present study is to utilize the technique of inverse problem to estimate the thermal properties and thickness of frost on the tube of evaporator.
This kind of problem is classfied as moving boundary problem as well as the inverse problem . In this paper the Conjugate Gradient Method is used for the numerical solution to the inverse geometry problem of identifying the unknown thickness of frost form temperature measurements, based on a boundary element method .
The use of boundary element method is suggested bye the basic nature of the inverse problem (to search an unknown domain, thus an unknown surface), because domain discretization is avoided. More specifically, the advantages gained by BEM-based algorithm is the ability to readily accommodate the changes in the unknown boundary shape as it evolves from its initial to its final shape.
|
|---|
