Newton’s iterative technique is commonly used in solving a system of non-linear equations. The advantage of using Newton’s method is that it gives local quadratic convergence leading to high computational efficiency. Specifically, Newton’s method has been applied to finite volume formulation for convective-diffusive transport processes. A direct solution method is adopted. Development of sparse direct solvers has significantly reduced the computation time of direct solution methods. Here UMFPACK (Unsymmetric Multi-Frontal method), has been used to solve the resulting linear system obtained from Newton’s step. A simple damping strategy is applied to ensure the global convergence of the system of equations during the first few iterations. The efficiency of this method is compared to that of Picard’s iterative procedure and the SIMPLE procedure for convective-diffusive transport processes. A modified Newton technique is also analyzed which lead to significant reduction in total CPU time.
- Heat Transfer Division and Electronic and Photonic Packaging Division
Application of Direct Newton’s Methods for Efficient Solution to Convective-Diffusive Transport Processes
- Views Icon Views
- Share Icon Share
- Search Site
Raju, MP. "Application of Direct Newton’s Methods for Efficient Solution to Convective-Diffusive Transport Processes." Proceedings of the ASME 2005 Summer Heat Transfer Conference collocated with the ASME 2005 Pacific Rim Technical Conference and Exhibition on Integration and Packaging of MEMS, NEMS, and Electronic Systems. Heat Transfer: Volume 3. San Francisco, California, USA. July 17–22, 2005. pp. 847-851. ASME. https://doi.org/10.1115/HT2005-72210
Download citation file: