Dear Prof Wolfgang,
Hi! I am successful in running the code and
it is without a doubt a nice addition to deal.ii code gallery. It
reproduces the expected result for 3d Beam case now.
Since I did not find any thermomechanical coupled example so far in deal.ii
therefore I want to contribute in the development and widening the scope of
this code by coupling it with heat equation of step-26.
For this purpose , taking your lectures as a reference, Initially for
testing purpose (i.e. before importing temperature solution from step-26) ,
I have added the thermal load in the *cell_rhs* as following:
* SymmetricTensor<2, dim> thermal_strain_qpoint;
/////////////////////////// assigning thermal strain as +ve temperature
change in whole body (i.e. all cells expanding equally)
//////////////////////// for
(unsigned int i=0; i<dim; ++i)
for (unsigned int j=0; j<dim; ++j)
if (i==j)
thermal_strain_qpoint[i][j] = (1-0.3) * 600e-6;//1e-3 ; // (1-nu) *
alpha * del T else
thermal_strain_qpoint[i][j] =
0*1e-3; // no shear thermal strain in case of homogenous material
properties
////////////////////////////////////////////////////////////////////////////////////////////////////////////*
* for (unsigned int j = 0; j < dofs_per_cell;
++j) cell_matrix(i, j) +=
(stress_phi_i *
fe_values[displacement].symmetric_gradient(j,
q_point) *
fe_values.JxW(q_point)); cell_rhs(i) +=
( (
stress_phi_i *
incremental_strain_tensor[q_point] )
- ( (
stress_strain_tensor *
fe_values[displacement].symmetric_gradient(i,
q_point)) * tmp_strain_tensor_qpoint )
+ (
fe_values[displacement].value(i,
q_point) * rhs_values_body_force )
+ (
fe_values[displacement].symmetric_gradient(i, q_point)
* stress_strain_tensor * thermal_strain_qpoint ) //
adding the thermal loading force produced from thermal
strains ) * fe_values.JxW(q_point);*
The solid geometry is in the form of rectangular plate (as shown in the
attachment where mesh is even finer in the example being tested) with zero
displacement (Dirchlet BC) at left edge and so far no mechanical traction
load is applied.
In the result the displacement result is qualitatively as per the
expectation i.e. the plate is expanding from other three ends but the
System residuals in Newton iterations are of the order of 1e+4 or even more.
I do not have much experience in under concentration mathematical approach
or I might be mistaking in writing the code for thermal loading force term
as per deal.ii standards (because the *rhs_values_body_force *term already
present in code is even bigger in numerical quantity than the thermal
loading term which is added by me but still it lets system converge for
small residuals of 1e-5 during solution) therefore I would be grateful to
you for any guiding response in my this course of development. Thank you!
On Monday, June 24, 2019 at 10:20:51 AM UTC+2, Muhammad Mashhood wrote:
>
> Thank you Prof Wolfgang for your possible support and concern. I will be
> waiting for the further support from the author to successfully run it and
> to understand the workflow correctly.
>
> On Saturday, June 22, 2019 at 1:51:10 AM UTC+2, Wolfgang Bangerth wrote:
>>
>>
>> Muhammad,
>> let me bring the original author of the program (Seyed Shahram Ghorashi)
>> into
>> the loop as well.
>>
>> > I ran the case of "Cantiliver_beam_3d" and it returned the results
>> after
>> > couple of minutes as mentioned in the attachement (where per time step,
>> > seemingly only one iteration was performed in Newton method and in all
>> > increamental load steps the result values are zero for both stress and
>> > displacement).
>>
>> Good question. I can confirm that that is what the program generates
>> (though
>> at least the dual solution is non-zero), but as with all of the code
>> gallery
>> programs, we just make them available for others to look at -- we don't
>> generally know what specifically they are doing, or whether they are
>> correct.
>> You will have to ask Seyed Shahram Ghorashi this question.
>>
>>
>> > Secondly I tried to run the case "Thick_tube_internal_pressure" which
>> first
>> > popped up the dialog message of :
>> > [...]
>> > Where as per my understanding of template based nature of deall.ii and
>> also
>> > this case is 2d so if I change the *dim = 2* from originally *dim = 3*
>> in main
>> > function of elastoplastic.cc
>>
>> Yes, this is the correct solution.
>>
>>
>> > then the following error comes during compilation :
>> > [...]
>>
>> Yes, I can confirm this as well. I will admit that I don't think this
>> ever worked.
>>
>> In the end, this is really the limitation of the code gallery. We see it
>> as a
>> way for our users to share codes that may be of use to others in the
>> community, but we do not endorse them or verify the quality of these
>> codes.
>> It's also quite possible that they *used* to work correctly with older
>> deal.II
>> versions, but don't any more -- I could imagine that being the situation
>> with
>> the first problem above.
>>
>> Let's hope that Seyed has something to add that help you!
>>
>> Best
>> WB
>>
>> --
>> ------------------------------------------------------------------------
>> Wolfgang Bangerth email: [email protected]
>> www: http://www.math.colostate.edu/~bangerth/
>>
>>
--
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 on the web visit
https://groups.google.com/d/msgid/dealii/eeefe5f2-0947-4e02-8f12-37d6dbe8f221%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
solution-0001-0000.eps
Description: PostScript document
