Dear All,

Thank you for your always support.

I am trying to serialize a Stokes solution in 2d ( dim for velocity +1 for 
pressure) to use it as a reference solution. Please stop me here if there 
is a better way to do this.

Unofortunatly, for some reason that I do not know, the saved_solution is 
all zeros or code breaks. 

To give more details I am solving using :
, fe(FE_Q<dim>(degree + 1), dim, FE_Q<dim>(degree), 1)
with BlockVector solution. This solution is not zeros ( I checked that).

Then I open a file and save in order the following:
<<triangulation
<<fe
<<dof_handler
<<solution
using 
boost::archive::text_iarchive
This finishes with no issues.  I even checked if the file is unreadable or 
empty, it is not.

When I deserialize to check the saved solution is matching my computed 
solution, I use same fe 
FESystem<dim> saved_fe(FE_Q<dim>(degree + 1), dim, FE_Q<dim>(degree), 1);
and deserialize in the following order
>>saved_triangulation
>>safe_fe
reint and distribute_dofe for the saved_dof_handler using the saved 
triangulation and fe.
>>saved_dof_handler
>>saved solution


After a deep investigation, it seems like " >>saved_dof_handler" is where 
the code beaks with error 
"----------------------------------------------------
Exception on processing: 
input stream error
Aborting!
---------------------------------------------------- "
which means that saved_dof_handler cannot be read form the saved file. 
However,  I checked after distributing dofs and  "saved_dof_handler" has 
the correct n_dofs().

In this code I was just checking, so dof_handler is not really needed here, 
but in the real code it is need to write a function out of the solution. 
For the sake of debugging, I removed  " >>saved_dof_handler". Removing it 
gives all zeros solution vector.  This is suspicious, is it connected to 
the saved_dof_handler?? 

I think I have checked many other things to find the source of the issue 
with no success. Conclusion, saved triangulation and saved fe seems to be 
correct.

Any one knows what is happening here? I would appreciate any guidance one 
what could be a source for the issue.

Best,
Najwa

-- 
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 Groups 
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/dealii/b2e37661-b458-4dbb-a849-52c6f1744ec2n%40googlegroups.com.

Reply via email to