Open AccessNUMERICAL INVESTIGATION OF THE POLYMER MELT FLOW IN INJECTION MOLDING BY USING ILU PRECONDITIONED GMRES
Author(s) -
Uğur Önsel Türk,
Ali Ecder
Publication year - 1999
Publication title -
mathematical modelling and analysis/mathematical modeling and analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.491
H-Index - 25
eISSN - 1648-3510
pISSN - 1392-6292
DOI - 10.3846/13926292.1999.9637122
Subject(s) - preconditioner , generalized minimal residual method , krylov subspace , mathematics , discretization , linear system , iterative method , rate of convergence , flow (mathematics) , computer science , mechanics , mathematical optimization , mathematical analysis , physics , geometry , channel (broadcasting) , computer network
The implementation of a modern preconditioned Newton‐Krylov solvers to the polymer melt flow in injection molding is the main focus of this paper. The viscoelastic and non‐isothermal characteristics of the transient polymer flow is simulated numerically and the highly non‐linear problem solved. This non‐linear behavior results from the combination of the dominant convective terms and the dependence of the polymer viscosity to the changing temperature and the shear rate. The governing non‐Newtonian fluid flow and energy equations with appropriate approximations are discretized by finite differencing. Elliptic Grid Generation technique is used to map physical domain to computational domain. The resulting non‐linear system is solved by using Newton's method. GMRES, one of the Krylov subspace methods, used as an iterative algorithm in order to solve the linear system at each non‐linear step. Incomplete LU preconditioner is used for better convergence. Numerical solution of polymer flow is presented to demonstrate that these methods are efficient and robust for solving such flow problems.