What is the physical meaning of the option "g(gradient)" in “# Boundary conditions for phase field in each direction”?
Choosing g(gradient), how does MICRESS assume the values of phase field parameter at imaginary nodes outside the domain?
g(gradient)
Re: g(gradient)
Dear mtoloui,
using a gradient boundary condition means extrapolating the field value to the first imaginary node outside the domain by using the gradient between the first and second neighbour of this node inside the domain. We allow the use of this boundary condition for all fields (concentration, temperature, phase-field) in order not to make any unnecessary restrictions for the user, even if we do not have a clear idea what it could be used for. Especially for the phase-field, it is hard to dream of a situation where the gradient condition may be needed, but who knows...
Bernd
using a gradient boundary condition means extrapolating the field value to the first imaginary node outside the domain by using the gradient between the first and second neighbour of this node inside the domain. We allow the use of this boundary condition for all fields (concentration, temperature, phase-field) in order not to make any unnecessary restrictions for the user, even if we do not have a clear idea what it could be used for. Especially for the phase-field, it is hard to dream of a situation where the gradient condition may be needed, but who knows...
Bernd
Re: g(gradient)
Sorry, I have to correct the last said: The gradient condition for phase-field is very useful for grain growth, because - if periodic is not possible for any reason - it should be the best choice for minimising the impact of the boundary condition on the grain structure!
Bernd
Bernd