Choosing the forcing terms in an inexact Newton method
SIAM Journal on Scientific Computing - Special issue on iterative methods in numerical linear algebra; selected papers from the Colorado conference
The Design and Use of Algorithms for Permuting Large Entries to the Diagonal of Sparse Matrices
SIAM Journal on Matrix Analysis and Applications
Iterative Methods for Sparse Linear Systems
Iterative Methods for Sparse Linear Systems
SIAM Journal on Scientific Computing
Hi-index | 31.45 |
Numerical aspects of computational modeling of chemical vapor deposition are discussed. Large sparse strongly nonlinear algebraic systems are to be solved per time step. For this, inexact Newton methods and preconditioned Krylov subspace methods are suitable. To ensure positivity of concentrations, we propose a novel approach, namely a projected inexact Newton method. Unlike the commonly used method of clipping, this conserves mass. Efficiency of several preconditionings is compared. Our numerical tests culminate in an unusually large computation, namely a three-dimensional case with 17 species and 26 reactions.