Hi Daniel,
The vectors you have in this equation P(u+h*v) =..., which of those have
constrains distributed and which zeroed?
If you assembly matrices with ConstraintMatrix.distribute_local_to_global()
the diagonal elements corresponding to constrained DoFs
will be dummy (and positive) just to m
Dear Stephen,
The first problem you are seeing is a bug in AlignedVector::push_back.
We merged a pull request last night,
https://github.com/dealii/dealii/pull/3993, which should fix this issue.
The problem was that we incorrectly invoked placement new with default
constructor rather than the copy
Daniel,
Hi all -- I'm writing a library that involves solving a nonlinear elliptic
PDE, which I'll write as f(u) = 0. There is an exact solution for this PDE for
a certain simplified geometry. To test everything, I check that my numerical
solution is tolerably close to the analytic solution, wi
Aycil,
Hi, I am a new user and I want to find out whether I can use deal.ii to solve
a 2D problem where there are two unknowns; a scalar function of x,y and a
scalar function of only x. I would like to initialize these as a vector and my
test functions will have the same structure.
Is it possi
On 02/22/2017 06:41 AM, Jiaqi ZHANG wrote:
Soluiton2 is fine. For Solution1, I checked the result with MATLAB, the
contour plot is right.
I also follow your instruction and tried data_out.build_patches(2).
Try build_patches(6) or build_patches(10).
As I stated, (i) what you output is not what
Hello,
I am solving a similar problem. What I do is to sum the contributions from
all the processors, and then use Utilities::MPI::Sum() to get the overall
force which I then write in a file on a master process,
For the contribution of a single cell, you have to integrate stress X
normal X are
Hi, I am a new user and I want to find out whether I can use deal.ii to
solve a 2D problem where there are two unknowns; a scalar function of x,y
and a scalar function of only x. I would like to initialize these as a
vector and my test functions will have the same structure.
Is it possible
Hi all -- I'm writing a library that involves solving a nonlinear elliptic
PDE, which I'll write as f(u) = 0. There is an exact solution for this PDE
for a certain simplified geometry. To test everything, I check that my
numerical solution is tolerably close to the analytic solution, with both
Hello,
I want to make an output of the resultant force existing on external
surfaces of a cube domain in a parallel code. what I did in the serial code
was that I integrated stress on external faces and as a result, I got one
scalar resultant force for the face to be written in a text file in e
Thank you Denis, that is what I am planing do to. Hope it will work out
finally.
Bests,
Zhenlin
在 2017年2月22日星期三 UTC-5上午3:53:26，Denis Davydov写道：
>
> you can always partition regular triangulation by
> GridTools::partition_triangulation (n_mpi_processes, triangulation);
> and adjust your assembly
Hi Franco,
Great, I'm glad that you managed to work it out!
Best,
Jean-Paul
On Wednesday, February 22, 2017 at 12:58:45 PM UTC+1, Franco Milicchio
wrote:
>
> Thanks guys, I've succeeded in making a full 3D example with your
> suggestions.
>
>
--
The deal.II project is located at http://www.d
Hi Zhenlin,
I concur with Denis - I regularly use the method that he's outlined in
conjunction with hp::DoFHandler.
Regards,
Jean-Paul
On Wednesday, February 22, 2017 at 9:53:26 AM UTC+1, Denis Davydov wrote:
>
> you can always partition regular triangulation by
> GridTools::partition_triangul
Hey Bruno,
Thanks for your reply, that makes sense to me now.
Best,
Jiaqi
在 2017年2月22日星期三 UTC-5上午9:13:29，Bruno Turcksin写道：
>
> Hi,
>
> while it looks strange I don't think there is any problem. In
> plottingBug-0002.png, Visit only knows the value at the four vertices of
> the domain and then
Hi,
while it looks strange I don't think there is any problem. In
plottingBug-0002.png, Visit only knows the value at the four vertices of
the domain and then it tries to "fill" the domain with a possible solution.
In figure1.png, you give Visit the values on 9 points and thus, it has more
inf
Thanks guys, I've succeeded in making a full 3D example with your
suggestions.
--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see
https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google G
you can always partition regular triangulation by
GridTools::partition_triangulation (n_mpi_processes, triangulation);
and adjust your assembly loops to
if (cell->subdomain_id == this_mpi_process)
to parallelize with MPI. The downside is that all your processes will now
own the **complete mesh**
16 matches
Mail list logo
