- Thank you received: 0
Forum Message
We have moved the forum to https://forum.ngsolve.org . This is an archived version of the topics until 05/05/23. All the topics were moved to the new forum and conversations can be continued there. This forum is just kept as legacy to not invalidate old links. If you want to continue a conversation just look for the topic in the new forum.
Notice
The forum is in read only mode.
Problem witch boundary conditions
4 years 4 months ago #2224
by LilBeng
Problem witch boundary conditions was created by LilBeng
Hello to all,
is there any way to enable the boundary conditions of gluing between bodies?
For example, if there are two boundaries, then the displacements u | Г1 = u | Г2 should be equal at these boundaries. Tell me please.
is there any way to enable the boundary conditions of gluing between bodies?
For example, if there are two boundaries, then the displacements u | Г1 = u | Г2 should be equal at these boundaries. Tell me please.
- mneunteufel
- Offline
- Premium Member
-
Less
More
- Thank you received: 59
4 years 4 months ago #2225
by mneunteufel
Replied by mneunteufel on topic Problem witch boundary conditions
Hi LilBeng,
if you have for example two Poisson equations decoupled like this
then you can either use a penalty formulation by adding
or introduce a hybrid facet variable (one additional FESpace).
Attached you'll find the full example.
Best,
Michael
if you have for example two Poisson equations decoupled like this
Code:
fes1 = H1(mesh, order=2, definedon= "left_dom", dirichlet="left|top|bottom")
fes2 = H1(mesh, order=2, definedon= "right_dom", dirichlet="right|top|bottom")
fes = FESpace( [fes1,fes2] )
(u1,u2),(v1,v2) = fes.TnT()
a = BilinearForm(fes)
a += 0.1*grad(u1)*grad(v1)*dx(definedon=mesh.Materials("left_dom"))
a += 0.15*grad(u2)*grad(v2)*dx(definedon=mesh.Materials("right_dom"))
a.Assemble()
f = LinearForm(fes)
f += 1*v1*dx(definedon=mesh.Materials("left_dom"))
f += 1*v2*dx(definedon=mesh.Materials("right_dom"))
f.Assemble()
then you can either use a penalty formulation by adding
Code:
a += 1e5*(u1-u2)*(v1-v2)*ds(definedon=mesh.Boundaries("interface"))
or introduce a hybrid facet variable (one additional FESpace).
Attached you'll find the full example.
Best,
Michael
Attachments:
4 years 4 months ago #2226
by LilBeng
Replied by LilBeng on topic Problem witch boundary conditions
Thanks for the quick response!
But what about in theory, when we have such a situation?
But what about in theory, when we have such a situation?
Code:
eo = SplineGeometry()
pnts = [(0, 0), (0.4999, 0), (0.4999, 1), (0, 1), (0.5, 0), (1, 0), (1, 1), (0.5, 1)]
pind = [ geo.AppendPoint(*pnt) for pnt in pnts ]
geo.Append(['line',pind[0],pind[1]],leftdomain=1,rightdomain=0,bc="bottom")
geo.Append(['line',pind[1],pind[2]],leftdomain=1,rightdomain=0,bc="interface")
geo.Append(['line',pind[2],pind[3]],leftdomain=1,rightdomain=0,bc="top")
geo.Append(['line',pind[3],pind[0]],leftdomain=1,rightdomain=0,bc="left")
geo.Append(['line',pind[4],pind[5]],leftdomain=2,rightdomain=0,bc="bottom")
geo.Append(['line',pind[5],pind[6]],leftdomain=2,rightdomain=0,bc="right")
geo.Append(['line',pind[6],pind[7]],leftdomain=2,rightdomain=0,bc="top")
geo.Append(['line',pind[7],pind[4]],leftdomain=2,rightdomain=0,bc="interface")
Attachments:
Time to create page: 0.120 seconds
