Hi
This example runs several linear solves because there is a nonlinear solver
below a continuation scheme. I figured that it would be possible to reuse the
preconditioner and see savings so I included the lines:
LinearSolver * linear_solver = system.get_linear_solver();
linear_solver->reuse_
Dear Libmesh developers and users,
I need to solve the following problem
1) Create a mesh
2) refine part of it for computing one 3D integral and deactivate
elements beyond that part
3) compute integral on the active elements
4) revert the mesh to the non-refined state
5) refine a different part of
On Thu, Nov 10, 2016 at 1:42 PM, Salazar De Troya, Miguel <
salazardet...@llnl.gov> wrote:
> I see. Do you have a reference or example on how this combination is put
> together?
>
systems_of_equations_ex2 and 3 solve the incompressible Navier-Stokes
equations using a Q2-Q1 mixed formulation on a
I see. Do you have a reference or example on how this combination is put
together?
Miguel
From: John Peterson mailto:jwpeter...@gmail.com>>
Date: Thursday, November 10, 2016 at 12:19 PM
To: Miguel Salazar mailto:salazardet...@llnl.gov>>
Cc: David Knezevic
mailto:david.kneze...@akselos.com>>,
"
On Thu, Nov 10, 2016 at 1:14 PM, Salazar De Troya, Miguel <
salazardet...@llnl.gov> wrote:
> That’s exactly what I was doing. I am playing with the example
> https://github.com/libMesh/libmesh/blob/master/examples/
> systems_of_equations/systems_of_equations_ex7/systems_of_equations_ex7.C
> where
That’s exactly what I was doing. I am playing with the example
https://github.com/libMesh/libmesh/blob/master/examples/systems_of_equations/systems_of_equations_ex7/systems_of_equations_ex7.C
where the mesh is built with HEX27, but the finite element is first order
lagrange. Why is there this mi
On Thu, Nov 10, 2016 at 3:01 PM, David Knezevic
wrote:
> On Thu, Nov 10, 2016 at 2:57 PM, Salazar De Troya, Miguel <
> salazardet...@llnl.gov> wrote:
>
>> Hello
>>
>> To implement the rigid body modes I need the node coordinates and the
>> dofs they are associated with, or vice versa, the dofs an
On Thu, Nov 10, 2016 at 2:57 PM, Salazar De Troya, Miguel <
salazardet...@llnl.gov> wrote:
> Hello
>
> To implement the rigid body modes I need the node coordinates and the dofs
> they are associated with, or vice versa, the dofs and a coordinate they are
> associated with. I can iterate over the
Hello
To implement the rigid body modes I need the node coordinates and the dofs they
are associated with, or vice versa, the dofs and a coordinate they are
associated with. I can iterate over the nodes to obtain their coordinates and
to obtain their dofs with DofObject::dof_number(), but this
On Tue, 8 Nov 2016, Salazar De Troya, Miguel wrote:
I want to attach the rigid body modes to the solver for AMG
coarsening. I saw there’s an interface in libMesh with
libMesh::NonlinearSolver
(https://libmesh.github.io/doxygen/classlibMesh_1_1NonlinearSolver.html#a2f0c2db92008f6e7c011861a649ec0
10 matches
Mail list logo
