Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2019-07-09 16:06:03

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Loi de comportement élasto-plastique[RESOLU][SOLVED]

Je veux comprendre pourquoi je ne obtiens pas le même état de déformation (et par consequence le même état de contrainte) si je utilise la commande ELAS or VMIS_ISOT_LINE avec des petits déformations (alors la limite de plasticité n'est pas atteint).

Ce n'est pas la même rélation lineaire jusqu' à la limite d'élasticité?

Merci beaucoup en avance!

Last edited by argi (2019-07-11 10:25:10)

Offline

#2 2019-07-10 04:01:35

Anirudh
Member
Registered: 2017-09-11
Posts: 243

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Hi,
Is it really the case, can you share some numbers?

Regards
Anirudh

Offline

#3 2019-07-10 10:14:16

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Hi Anirudh,

I send you the script.comm, 3couche.med (mesh) and export file, together with the results printed in the resELAS.med file. If we open this file with ParVis (Salome), we obtain the following results:

DX_MAX=0   DX_MIN=-4.2 e-05
DY_MAX=8.6 e-07   DY_MIN=-2.6 e-06
EPSI_NOEU  Magnitude MAX= 2.1 e-03       MIN=2.3e-04
SIGM_NOEU  Magnitude MAX= 2.2 e+08       MIN= 1.3 e+05

while replacing COMPORTEMENT=ELAS by COMPORTEMENT= VMIS_ISOT_LINE (the limite of elasticity is set to be SY=2.5e+08 by ECRO_LINE in the .comm file) we obtain the results saved in resVMIS_ISOT_LINE.med file:

DX_MAX=2 e-20   DX_MIN=-2.4 e-38
DY_MAX=1.6 e-22   DY_MIN=-8.5 e-22
EPSI_NOEU  Magnitude MAX= 2.2 e-03       MIN=2.3e-20
SIGM_NOEU  Magnitude MAX= 2.5 e-07       MIN= 1.1 e-09

Any hints to understand this results?

Thank you in advance!


Attachments:
script.comm, Size: 3.15 KiB, Downloads: 31

Offline

#4 2019-07-10 10:15:35

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

I have attached the mesh file, I can not upload the results as they are too big.
If you are interested, I can send you by email!

Last edited by argi (2019-07-10 10:29:20)


Attachments:
3couche.med, Size: 759.97 KiB, Downloads: 25

Offline

#5 2019-07-10 10:39:39

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

The expot file (I have not found a way to upload different files at the same time)


Attachments:
export, Size: 464 B, Downloads: 25

Offline

#6 2019-07-10 11:43:14

Anirudh
Member
Registered: 2017-09-11
Posts: 243

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Hello,
Just some quick thoughts.
The SY(Elastic limit) of your second. comm is almost similar to the max stress from your first. comm file. That means stresses reach plasticity level and you can get some plastic strain in second case under VMIS_ISOT_LINE.
Try again with very high SY and a positive D_SIGM_EPSI. A value of zero D_SIGM_EPSI indicates perfect plastic strain which I think is unusual in practice.

Regards
Anirudh

Offline

#7 2019-07-10 16:26:49

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,569

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

hello

@argi
i do not understand what you are trying to do

1- you have an enormous (to my eyes) number of nodes in the mesh, but the problem may not come from that

2- you are performing a non linear analysis with a load that is constant,

3- you are calculating with one single step over the range of the problem

in these conditions you are going to have trouble to catch any non linearity if code_aster wants to solve the problem

@Anirudh
many construction codes recommend (if not impose) to use a flat curve beyond the elastic limit i.e. ET=0 or  D_SIGM_EPSI=0.0 = 0

jean pierre aubry

Offline

#8 2019-07-10 18:03:49

Anirudh
Member
Registered: 2017-09-11
Posts: 243

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Hi Jean Pierre,
If D_SIGM_EPSI is 0 then strain hardening effect will not be present in the material after unloading. I don't know if such a model is valid physically because some materials especially ductile metals exhibit a higher yield strength after permanent deformation.

Anirudh

Offline

#9 2019-07-11 08:51:09

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Hello @Jeanpierre,

The main goal is to compute plastique strain from an initial temperature difference.

1) I have tested reducing the nodes, and using incremental time steps (see 3), and I obtain the same results, deplacement, strain and stress almost null.

2) The second step (after running this one) would be to start another elasto-plastique simulation taking as an initial state the stress that we obtain from this simulation. This is why I use non-linear simulation, because the linear one does not accept an initial state. In addition, the main goal is to compute plastique strain, which is imposible with the linear simulation (there in no COMPORTEMENT option, nor EPSI_NOEU using CALC_CHAMP).

3) Incremental time step
listr = DEFI_LIST_REEL(DEBUT=-1.0,
                       INTERVALLE=(_F(JUSQU_A=0.0,
                                      PAS=1.0),
                                   _F(JUSQU_A=1.0,
                                      NOMBRE=10)))

times = DEFI_LIST_INST(DEFI_LIST=_F(LIST_INST=listr))

But all propositions are welcome! Do you have more hints or ideas?

Thank you very much for your time!

Last edited by argi (2019-07-11 09:04:04)


Attachments:
3couche_BIG.med, Size: 62.95 KiB, Downloads: 36

Offline

#10 2019-07-11 09:03:54

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Hi @Anirudh,

The main goal, as I have written to Jean Pierre, is to compute the plastique strain.
Anyway, I have try with bigger yield strength (SY) and non null D_SIGM_EPSI, and still it does not work (meaning that the displacement, strain and stress fields are almost empty).

So more ideas? All your thoughts are welcome as I am quite blocked!

Thank you very much for your time!

Offline

#11 2019-07-11 09:44:46

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,569

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

attached is a corrected file that does the job with the same results in both case
temperature decreases from INST=-1 to 0 and remains constant until 1

it would be quite long to describe here the modifications i made
the code is commented


Attachments:
script.comm, Size: 3.84 KiB, Downloads: 39

Offline

#12 2019-07-11 10:20:19

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

Thank you very muchhh!!!!
Merci beaucoupp!!
Miless de graciasss!!
Eskerrik asko!!! smile smile

So the key is to compute two temperature fields and the create a evol_varc variable with CREA_RESU, to create incremental temperature field, isn't it?
Merci beaucoup un autre fois Jean Pierre!! smile smile

Offline

#13 2019-07-11 13:48:13

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,569

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

So the key is to compute two temperature fields and the create a evol_varc variable with CREA_RESU, to create incremental temperature field, isn't it?

yes indeed
as a non linear analysis with a constant load does not mean much

do not forget to intrduce yourself in the ad hoc section of the forum

Offline

#14 2019-07-11 15:54:07

argi
Member
From: Bidart FRANCE
Registered: 2019-01-10
Posts: 12

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

what is "ad hoc" section? Do you mean to submit a post introducing myself?

Offline

#15 2019-07-11 17:00:38

jeanpierreaubry
Guru
From: nantes (france)
Registered: 2009-03-12
Posts: 3,569

Re: Loi de comportement élasto-plastique[RESOLU][SOLVED]

there is section named "Introduce yourself / Présentez vous" on the master page of the forum

Offline