Include Stress as initial condition

1

Hello,

I want to check the behavior of a plate with an initial stress, but the result is not what I expect.

The idea is to apply a uniform pressure on a plate fixed on the right corner during a dynamic calculation.

config
config2490Ă—1415 21.5 KB

configopen
configopen2490Ă—1415 33 KB

Then I save the new deformed mesh and the stress at the last iteration, I run a new simulation with this mesh and this stress as initial condition but without any pressure.

If I do this new simulation in STATIC with the NLGEOM solver,the plate almost goes back to the initial state but there is still a gap about the value of displacement.

static
static2490Ă—1415 27.5 KB

When I check the stress in the _main.dat, strangely the stress in the element change a little but doesn’t tend to zero during the calculation.

If I do the same in DYNAMIC, the time step decrease and the plate doesn’t go back at all to its original flat shape (I stopped the calculation after 1 day).

dyna
dyna2490Ă—1415 46.2 KB

For me, in both cases the plate should go back naturally to a the original shape in order to cancel the initial stress?

For information, I apply the initial stress by using “*INCLUDE, INPUT=add_stress.inp” in my _main.inp with add_stress.inp the file containing the stress during the last iteration of the first calculation.

I appreciate any hints

2

Hi
I have tried this before. The most important hint I can give is telling you to check the stress types of output and input files of Calculix.
As I recall,
the output stress in dat file is SXX, SYY, SZZ, SXY, SXZ, SYZ
however,
the input of ccx should be SXX, SYY, SZZ, SXY, SYZ, SXZ

Also, I have tried the same thin with abaqus by making an analysis with two steps:
I have fixed one side of bar before the step initialization,
then applied pressure at first step
and then to removed the pressure by typing *DLOAD, OP=NEW
I know that you have two analyses, one of which inherits the conditions from the other.
If you can share the inp file here I can check the conditions.

PS: You can run the same analysis using *INITIAL CONDITIONS, TYPE= DISPLACEMENT and exporting the displacement field of each node to the following lines.

1 Like
3

Hello

Ok, now I can go back to the initial state by using “*DLOAD, OP=NEW” in a multi step run.

About start a simulation with an initial stress as condition, for the moment I still don’t have very good results in static even by switching SXZ and SYZ, I will check the conditions.

Thank you