I've added the "no_requirements" in FiniteElementDomination namespace.
There is still a problem with FE_System.
I've tested a simplier case to check it, i.e. a FE_System with only one
element:
fe_1 ( FE_Q<dim>(1), 1 ),
fe_2 ( FE_Nothing<dim>(), 1 ),
fe_collection.push_back (fe_1);
fe_collection.push_back (fe_2);
I got the following error message:
Program received signal SIGSEGV, Segmentation fault.
0x00007ffff71ad244 in dealii::FESystem<2, 2>::get_data (this=0x685df0,
flags_=14395, mapping=..., quadrature=...)
at
deal.II/source/fe/fe_system.cc:726
726 Assert (!(base_fe_data->update_each & update_hessians),
in file "fe_system.cc" before line 726 there is the comment
// make sure that *we* compute
// second derivatives, base
// elements should not do it
FE_Nothing from "update_once" and "update_each" gives a "update_default",
i.e. no update.
It seems that the update mechanism in a FE_System with a FE_Nothing has to
be modified.
Suggestions?
Best
tc
-------------------------------------------------
On Sun, 6 Feb 2011, Wolfgang Bangerth wrote:
I'll implement it and I'll test the case
Q1 Nothing
Nothing Q1
for the case
Q1 Nothing
Nothing Q2
Thanks. Let me know if you run into trouble!
I'm not sure (I've to read the code) what I need for imposing the fluxes
on both sides having Q1 on one and Q2 on the other side.
That's the usual hp-case that is already handled in
DoFTools::make_hanging_node_constraints.
Best
W.
-------------------------------------------------------------------------
Wolfgang Bangerth email: [email protected]
www: http://www.math.tamu.edu/~bangerth/
_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
