Deprecating MeshData would require the methods of UNVIO etc. to be
rewritten. I'm glad I'm not the one to figure out the classes and methods
going to be involved when transferring from a file to an explicitsystem.
As for me, I am using my own function to transfer from my data structures
to the mes
On Tue, 7 Aug 2012, Onur Solmaz wrote:
> I'm using Tet4 elements.
> How do I compile libmesh in debug mode? I couldn't figure out.
Using a Bourne-type shell, "export METHOD=dbg" before you compile
libMesh or run any examples; or just replace "make" with "METHOD=dbg
make".
> I'm using the most r
I'm using Tet4 elements.
How do I compile libmesh in debug mode? I couldn't figure out.
I'm using the most recent svn build by the way.
Onur
2012/8/7 Roy Stogner
>
> On Tue, 7 Aug 2012, Onur Solmaz wrote:
>
> It would be great if there were an example on that ExplicitSystem.
>>
>
> *Definitely
On Tue, 7 Aug 2012, Onur Solmaz wrote:
> It would be great if there were an example on that ExplicitSystem.
*Definitely*. Trouble is finding someone with both the knowhow and
the time to sit back and write it up.
> I also had another weird problem. Would there be a reason that the
> number of
It would be great if there were an example on that ExplicitSystem.
I also had another weird problem. Would there be a reason that the number
of dofs won't be equal to the number of nodes?
It is weird because my program works perfectly fine with my first input
file (n_dofs=n_nodes), but it crashed
On Tue, 7 Aug 2012, Jens Lohne Eftang wrote:
> Is there a way to remove a vector from a system, by name, when this
> vector has already been added to a system with add_vector("my_vector") ?
Currently no. But System::remove_vector() implementations shouldn't
be too hard to add; if you want to se
On Tue, 7 Aug 2012, Onur Solmaz wrote:
Thanks for the detailed response!
No trouble. Try to hit "Reply to All" rather than just "Reply" to
libmesh messages, though - I like making sure these threads get
archived so Google can help others with similar issues in the future.
Calling Lib
One way to solve this problem while ensuring type safety (or at least get
rid of the need for dynamic casts)
is to attach all callbacks directly to the system, which wouldn't know
anything about the solver.
The solver then would have a reference to NonlinearImplicitSystem and
forward all PETSc call
Is there a way to remove a vector from a system, by name, when this
vector has already been added to a system with add_vector("my_vector") ?
Jens
--
Live Security Virtual Conference
Exclusive live event will cover all t
On Tue, 7 Aug 2012, Ataollah Mesgarnejad wrote:
>> TransientNonlinearImplicitSystem& realsys =
>> libmesh_cast_ptr(sys);
>>
>
> This didn't work and gave a not matching function error. However:
>
> TransientNonlinearImplicitSystem& realsys =
> dynamic_cast(sys);
>
> worked.
My mistake - I mean
On Tue, 7 Aug 2012, Onur Solmaz wrote:
Calling LibmeshInit in my main function gives me segfaults for obvious reasons.
I'm afraid not obvious to me - all our examples and nearly all my
application codes create a LibMeshInit object in main().
And I can't allocate from the heap:
LibMeshInit *
On Tue, 7 Aug 2012, Ataollah Mesgarnejad wrote:
> I'm trying to impose some irreversibility bounds on my system that requires
> updating bounds based on the previous time step solution. Is there a trick to
> do this ?
>
> I tried to pass:
>
> void FormBounds(NumericVector& XL,NumericVector&
>
Dmitry,
I'm trying to impose some irreversibility bounds on my system that requires
updating bounds based on the previous time step solution. Is there a trick to
do this ?
I tried to pass:
void FormBounds(NumericVector& XL,NumericVector&
XU,TransientNonlinearImplicitSystem& sys);
as "system
13 matches
Mail list logo
