Inverse Design of Solidification Processes
Kei Okamoto, Ben Q. Li
Abstract
Kei Okamoto, Ben Q. Li
Abstract
An inverse algorithm is developed for the design of the solidification processing systems. The algorithm entails the use of the Tikhonov regularization method, along with an appropriately selected regularization parameter. Both the direct solution of moving boundary problems and the inverse design formulation are presented, along with the L-curve to select an optimal regularization parameter for inverse design calculations. The design algorithm is applied to determine the appropriate boundary heat flux distribution in order to obtain a unidirectional solidification front in a 2-D cavity by eliminating the effect of natural convection. Inverse calculation is also performed for the case in which the solid-liquid interface is prescribed to vary linearly. The L-curve based regularization method is found to be reasonably accurate for the purpose of designing solidification processing systems.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
An inverse algorithm is developed for the design of the solidification processing systems. The algorithm entails the use of the Tikhonov regularization method, along with an appropriately selected regularization parameter. Both the direct solution of moving boundary problems and the inverse design formulation are presented, along with the L-curve to select an optimal regularization parameter for inverse design calculations. The design algorithm is applied to determine the appropriate boundary heat flux distribution in order to obtain a unidirectional solidification front in a 2-D cavity by eliminating the effect of natural convection. Inverse calculation is also performed for the case in which the solid-liquid interface is prescribed to vary linearly. The L-curve based regularization method is found to be reasonably accurate for the purpose of designing solidification processing systems.
Key concepts: Tikhonov regularization, Inverse problem, Regularization (linguistics), Inverse, Inverse method, Boundary value problem, Heat flux, Applied mathematics