Elasticity probelm: same matrices - different solution vector.

252 views
Skip to first unread message

Мария Бронзова

unread,
Oct 24, 2022, 3:16:01 PM10/24/22
to deal.II User Group
Dear all,

I have been trying to implement orthotropic elasticity (similar as in step-18) and in order to verify my results I decided to set all E and G modulus values as in an isotropic case and to compare the outcome with the one for the same isotropic elastic problem but with another formulation (as in step-46 for the structure part). 

I created the simplest 1 cell models and compared the matrices in both cases (both LHS and RHS), they are absolutely identical. But weirdly enough I receive different solution vectors. 

I have been struggling trying to identify the source of the problem. If you could please take a look at my small scripts - I would be very grateful..(you can find them attached).

Thank you a lot in advance!

Kind regards,
Mariia


orth_1_cell_no_freq
elast_1_cell_no_freq

Wolfgang Bangerth

unread,
Oct 24, 2022, 5:14:11 PM10/24/22
to dea...@googlegroups.com
On 10/24/22 13:16, Мария Бронзова wrote:
> *** Caution: EXTERNAL Sender ***
>
> Dear all,
>
> I have been trying to implement orthotropic elasticity (similar as in
> step-18) and in order to verify my results I decided to set all E and G
> modulus values as in an isotropic case and to compare the outcome with
> the one for the same isotropic elastic problem but with another
> formulation (as in step-46 for the structure part).
>
> I created the simplest 1 cell models and compared the matrices in both
> cases (both LHS and RHS), they are absolutely identical. But weirdly
> enough I receive different solution vectors.

Maria:
That is of course a contradiction. If you have the same linear system,
and you solve it the same way, you should get the same solution.

We are generally happy to help with advice on this forum, but we can't
debug other people's code. But I think you have everything you need to
debug your codes yourself. Just output the elements of the matrix and
rhs of the linear system for both codes. Make sure they are the exact
same. If they are not, then you know that the problem lies with the
assembly. If they are, then solve the linear system with the exact same
solver and output the solution element by element. It should be the
exact same. If it is not, you know that the problem lies with what must
be different ways you are solving the linear system. If it is, then the
problem lies in the part of the code that does something with the
solution later on.

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/
Reply all
Reply to author
Forward
0 new messages