Hello,
I have a Coefficient form PDE with Dirichlet boundary conditions with two dependent variables. The dependent variables range -5<v1<5 for v1 and 250<v2<600 for v2. The model converges quickly and gives expected results when the diffusion coefs are f(v2) [ie f(v2) = av2 + b] and the source term is a f(v1v2,v1^2).
When I try to make the source term f(v1^2,v2^3) or the diffusion coef f(v2^5) then I cannot get the computation to converge. v1 and v2 are well defined real numbers throughout the range of interest. Is there some kind of constraint I can implement to allow convergence?
I do not know how to create a domain for v1 and v2 to restrict them to possible values but suspect that that might allow for a solution.
I have tried different solvers and mesh sizes.
Absorption coef, mass coef, damping coef, cons flux coef, conv coef and cons flux source are all zero. The solution is stationary time invariant.
I have a Coefficient form PDE with Dirichlet boundary conditions with two dependent variables. The dependent variables range -5<v1<5 for v1 and 250<v2<600 for v2. The model converges quickly and gives expected results when the diffusion coefs are f(v2) [ie f(v2) = av2 + b] and the source term is a f(v1v2,v1^2).
When I try to make the source term f(v1^2,v2^3) or the diffusion coef f(v2^5) then I cannot get the computation to converge. v1 and v2 are well defined real numbers throughout the range of interest. Is there some kind of constraint I can implement to allow convergence?
I do not know how to create a domain for v1 and v2 to restrict them to possible values but suspect that that might allow for a solution.
I have tried different solvers and mesh sizes.
Absorption coef, mass coef, damping coef, cons flux coef, conv coef and cons flux source are all zero. The solution is stationary time invariant.