Connecting multiple comm files

sikumegalifter

Dear all,
I am asking for your help to solve a problem with code aster.
I am asking if it is possible to write two or more comm files for the same model that are executed sequentially.
For example, in the first comm file I enter the model data and run the static analysis. In the second comm file, I use the results of the static analysis to calculate the seismic masses, add them to the model and run the modal analysis.
See the flowchart below.

I would like to have these operations done automatically because, as you can imagine, I have a considerable amount of data to deal with.
many thanks

AnatoleCG

Hello, you can find your answer here: https://forum.code-aster.org/public/d/27633-how-to-use-multiple-comm-for-a-calculation

sikumegalifter

AnatoleCG
Good morning
I have read the post you suggested but I cannot reproduce it in my case.
Please can you send me an example you made, both the comm file and the export?
There is probably something I don't quite understand how to do.
Thank you very much

AnatoleCG

Here's an example of your export files it will first do caract then statique then modale.
For caract you have to start your code with DEBUT() and end with FIN() for the rest you have to start with POURSUITE() and end with FIN(). If it is not clear i will create an easy to understand template

you can have an example in this file : https://file.io/N63hBbTj9Gsj

k_zurawski

If you run singularity container in terminal you may start ASTK from terminal. Write astk and it should start. Then you may create export file.

sikumegalifter

Thank you all for your help.
I have one more question.
Is it possible and how to run comm file driven by conditional choice?
For example, after a conditional choice I run comm2 file instead of comm3 file. See the flowchart below.
Many thanks to all

AnatoleCG

sikumegalifter In my opinion the easier way to do that is to fusionate your comm 2 and comm 3 and use condition with python. Or I never tried but you can try to use a variable at the end of comm 1 and make it pass or not your comm 2 or 3 in a linear way. As an example, you write a variable A in the end of your comm file. A can have 1 or 2 depending of the results in comm 1. You go to comm 2, if A=1 you do comm 2 and if it is 2 you pass comm 2 and go to comm 3 directly. You make the same with comm 3 for A=2. I don't know if it is clear enough

sikumegalifter

Thank you for your answer.
Is there a command I can write at the end of the Comm1 file that will allow continue execution with Comm2?

For example
If a = 0
execute file comm2
Else
execute file comm3
end

How can I translate "execute file comm2" into Code Aster language?
Thank you very much!

AnatoleCG

sikumegalifter I don't think so i would say it's more

comm1:
DEBUT()
execute the program
a =0 or 1
FIN()

then comm 2:
POURSUITE()
if a = 0:
execute the program
else:
pass
FIN()

then comm 3:
POURSUITE()
if a = 1:
execute the program
else:
pass
FIN()

sikumegalifter

Good morning everyone,
I am once again asking for your help.
I need to use the results of the static analysis (the nodal forces) calculated in the comm1 file, multiply them by a constant, and enter them as a load in the comm2 model.

The code I wrote is as follows, in which I do static analysis and write the nodal forces.
How do I multiply the nodal forces by a constant?
How do I enter these forces as loads in a new model?

thank you

_ Analysis G1 ____

resG1 = MECA_STATIQUE( MODELE=model,
CARA_ELEM=cara,
CHAM_MATER=mater,
SOLVEUR = F(METHODE = "MUMPS",
STOP_SINGULIER ="OUI",
RENUM ="METIS",
NPREC = 8,
ELIM_LAGR = "NON",),
EXCIT=(F(CHARGE=constr),
_F(CHARGE=G1),
))

Extracting nodal forces __

resG1=CALC_CHAMP(reuse=resG1,
FORCE=('FORC_NODA',),
RESULTAT=resG1)

_ Creating tables nodal forces ____

tFNG1=CREA_TABLE(RESU=_F(GROUP_MA=('Tutti',),
NOM_CHAM='FORC_NODA',
RESULTAT=resG1,
TOUT_CMP='OUI'),
TITRE='NODAL FORCES G1')

sikumegalifter

Good morning everyone,
To try and solve the problem, I used the command POST_RELEVE_T
and then converted the data to Python with .EXTR_TABLE()
At this point the following problems remain:
1- I need to multiply the forces DX,DY,DZ by a constant.
2- Apply the forces calculated in step 1 to the same structure (the nodes do not change).
3- Repeat the static analysis

Can you help me?
Thank you.

_ Creating tables nodal forces ____

tFNG1bis= POST_RELEVE_T(
ACTION=_F(
INTITULE='Forze nodali',
RESULTAT=resG1,
NOM_CHAM='FORC_NODA',
GROUP_NO='Tutti',
OPERATION='EXTRACTION',
TOUT_CMP='OUI' # Estrai tutte le componenti delle forze nodali
)
)

___________ Convert the data in Python

tab2 = tFNG1bis.EXTR_TABLE()

jeanpierreaubry

hello

i really do not understand what you are trying to fiddle with
is your ultimate goal doing a seismic analysis?

jean pierre aubry

sikumegalifter

Good morning Dr Aubry,
I need to do a seismic analysis of a structure as I told you in the other post.
To simplify the problem, the steps I am following are as follows:
1- I have a structure with applied loads
2- I perform a static analysis
3- I extract the forces (FORC_NODA) at each node of the mesh
4- I convert the forces calculated in step 3 into masses (FORC_NODA/9.81)
5- I apply these masses to the structure using POI1 elements DIS_T and M_T_D_N (as you suggest).
6- I carry out a modal analysis

I can do STEP 1,2,3 and 6.
I can do STEP 6 by manually entering the masses on the nodes, but on a small test model (10 nodes).
Since my structure has thousands of nodes, it seems obvious to me to automate this STEP.
To do STEP 4 and 5, I was thinking of using the Python language written in the comm file.
My idea was to do a FOR loop on vectors to read the node forces calculated in STEP 3, convert them to masses and apply them as masses to the same nodes using M_T_D_N.
If there are simpler methods I will gladly accept them.

I hope I have clarified my questions.
thanks

Sound_Spinning

sikumegalifter
Modal analysis doesn't have loads, you need to think of mass & stiffness only with the correct constraints, and you are done.
The loads that matter in your statics cases are the sum at all constraints, that IS the net force/moment vectors travelling thru your model, and you need to care about.
If what you mean is that you need a distributed mass over a surface or a node set, there is a single command to do that for the modal.
I recommend that you do not go the manual way for each point-mass, there are other (simpler) ways.

jeanpierreaubry

i wish you good luck with your way of doing

but it is puzzling me

as you know the force you apply in step 1, you know the mass that produces that force
why do not you apply the mass at this stage

you are certainly not the first one to make an analysis with thousands of nodes
and added mass
and i have never seen being done your way

i am not even sure that a simple FORC_NODA computation is meaningful of the applied load,
it depends of the domain of computation and the boundary condition

sikumegalifter

Good morning,
thank you all for your answers, they are helping me a lot!!!
I have another question on stage management.

I am simulating a construction by stages, in the COMM1 file I have only one floor of the building (e.g. 2 pillars and 1 beam), then I want to add in the COMM2 file two more pillars and another beam to simulate the construction of a second floor.

In the COMM1 file I have defined the model in this way:
model = AFFE_MODELE(AFFE=(
_F(GROUP_MA=('P_1'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('P_2'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('T_1'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
),
MAILLAGE=mesh)

In the COMM2 file I want to add more parameters to model, how do I do this?
Now I am going to rewrite a new variable (model2), but I think there is an easier way, can you help me?

model2 = AFFE_MODELE(AFFE=(
_F(GROUP_MA=('P_1'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('P_2'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('T_1'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('P_3'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('P_4'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
_F(GROUP_MA=('T_2'),
MODELISATION = 'POU_D_T',
PHENOMENE='MECANIQUE'),
),
MAILLAGE=mesh)

Many thanks
Greetings