Here's an example. It's trivial, but it just demonstrates solving a
variable on one grid and then using that variable to set the value of
the diffusion coefficient for an equation on a different grid.
Obviously, if the domains don't overlap then you need to take care of
the mapping and interpolation between the grids.
~~~
from fipy import CellVariable, Grid1D, DiffusionTerm, Viewer, TransientTerm
nx = 100
dt = 1e-3
Lx = 1.0
steps = 1000
mesh0 = Grid1D(nx=nx, Lx=Lx)
mesh1 = Grid1D(nx=nx, Lx=Lx)
v0 = CellVariable(mesh=mesh0, value=0., hasOld=True)
v0.constrain(1e-1, where=mesh0.facesLeft)
v0.constrain(10., where=mesh0.facesRight)
v1 = CellVariable(mesh=mesh1, value=0., hasOld=True)
v1.constrain(0., where=mesh1.facesLeft)
diff_coeff = CellVariable(mesh=mesh1, value=0.)
eqn0 = TransientTerm() == DiffusionTerm()
eqn1 = TransientTerm() == DiffusionTerm(diff_coeff) + 1.
viewer = Viewer(v1)
for i in range(steps):
v0.updateOld()
v1.updateOld()
eqn0.solve(v0, dt=dt)
diff_coeff[:] = v0
eqn1.solve(v1, dt=dt)
viewer.plot()
input("stopped")
On Thu, Apr 18, 2024 at 3:31 AM Nils Winkler <[email protected]> wrote:
>
> Oh great. Do you happen to have example code? I am not sure how to properly
> do that.
>
> Daniel Wheeler schrieb am Mittwoch, 17. April 2024 um 16:19:16 UTC+2:
>>
>> On Wed, Apr 17, 2024 at 9:33 AM Nils Winkler <[email protected]> wrote:
>> >
>> > Hi,
>> >
>> > I am using fipy to deal with a system of three coupled nonlinear partial
>> > differential equations. Two dependent parameters have Naumann BCs and one
>> > periodic BCs. Is it possible to implement these at the same time?
>> >
>> > Using only a periodic mesh with internal BCs for the Naumann BC does not
>> > work because the respective two dependent parameters are not continuous at
>> > the boundary.
>> > Can one somehow combine a periodic with a normal mesh? Or is it possible
>> > to use a normal mesh and impose periodic boundary conditions internally?
>>
>> Yes, you can do that. You'll have to define the dependent variables
>> across both the domains and map the variable values over at each sweep
>> / timestep so there will be some overhead.
>>
>> --
>> Daniel Wheeler
--
Daniel Wheeler
--
To unsubscribe from this group, send email to [email protected]
View this message at https://list.nist.gov/fipy
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
