Hello all,
I was wondering if there is plane strain element (generalized) in dealii,
i.e. the domain will be 2D whereas
the displacement will have all three components and the stress and strain
tensors will be 3x3. In my case
it would be similar to step 44 (but compressible) with finite strain
Hi friends,
I am parallelizing a code similar to step-44 in which it is possible to use
either an iterative solver SolverCG or a direct solver,
SparseDirectUMFPACK. I have used the latter in the non-parallel code and
works great.
Using iterative solvers like SolverCG I have problem in convergen
Tulio,
If you want to prescribe Dirichlet boundary conditions on a part of the
boundary of your domain, you can just set some boundary_id to the
respective cell boundary faces[1] and call
VectorTools::interpolate_boundary_values[2]. The boundary_ids are inherited
when you refine the mesh.
If h
H can be hydraulic head and K is the hydraulic conductivity.
Or H can be temperature with k being diffusion coefficient.
The boundary here consists of only flux in and flux out condition. So only
neumann boundary condition is given here.
On Monday, October 24, 2016 at 11:32:35 PM UTC+2, Danie
Hi Jean-Paul,
The last journal file that you sent me run successfully. I was able to read
the file that it generated using GridIn::read_ucd.
By the way, it would be more clear if the suffix of the generated file will
be "ucd" and not "inp" which is the suffix that is affiliated with abaqus
file
Retired Replicant,
What are you trying to model? What should H represent?
In general it should not be a problem to mix Dirichlet and Neumann boundary
conditions,
but this depends on what are you trying to do.
Best,
Daniel
--
The deal.II project is located at http://www.dealii.org/
For mailing
Hi Oded,
Hmm... I didn't have much luck with the script either. That was unexpected!
I've dug up an older version from many year ago and attached it. Please
give it a go! I know that this one doesn't output material IDs, but I'll
try to see if I can refactor it so that it does.
Regards,
J-P
O
> It would be appreciated if you could give me any clue to resolve the error.
You need a ghosted vector as the parameter in MappingQEulerian. Is
that the case for you?
See
https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_dealii_dealii_blob_master_tests_mappings_mapping-5Fq-5Feuler
Hi Jean-Paul,
I tried to run the python script that you noted, but got various error
messages.
Were you able to run it successfully on my mesh?
Best,
Oded
On Monday, October 24, 2016 at 11:05:36 AM UTC-4, Jean-Paul Pelteret wrote:
>
> Dear Oded,
>
> The error message from the development versi
Dear Oded,
The error message from the development version is more useful.
Unfortunately there's a bug somewhere in the routine that appears from
time-to-time on more complex meshes that I have not been able to track as
of yet. This is clearly one of those times... For now you could consider
us
Is this a correct boundary condition ?
equation is : div(k(x,y)del(H)) = q(x,y)
div is divergence, del is the usual gradient operator.
now, the domain is a rectangular shape in 2d.
The top and bottom edge are no flow boundary condition.
the left edge is a constant flux in (Qin)
the right edge
Hi, taking advantage of this topic, i'd like to ask a related question.
In a 2D mesh, suppose i want to enforce a dirichlet boundary for which i
have its nodal values. Edge's values could off course be obtained trough
its nodes interpolation.
In this context, I guess i could use an approach lik
12 matches
Mail list logo
