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

You are not logged in.

#1 2018-10-02 14:20:48

bhattarai
Member
Registered: 2014-01-31
Posts: 124

Chained calculations with two STAT_NON_LINEs

Hi all,

I would like to chain 2 STAT_NON_LINE calculations one after another as shown in attached lift.pdf file. I have previously done similar calculations with different (fixing the displaced geometry) boundary conditions at the end of first step. Following the same idea, for different boundary condition as described by following snippet brings the structure back in its initial state at the beginning of the 2nd STAT_NON_LINE:

cine1=AFFE_CHAR_MECA(MODELE=FEMNolin,
                     DDL_IMPO=(_F(GROUP_MA=('FIX',),
                                  DX=0,
                                  DY=0,
			          DZ=0,),
                               _F(GROUP_MA='LIFT',      # lifting the top of the column
				  DX=0.0,
                                  DY=0.0,
				  DZ=20.0,),),);

cine2=AFFE_CHAR_MECA(MODELE=FEMNolin,
                     DDL_IMPO=(_F(GROUP_MA=('FIX',),
                                  DX=0,
                                  DY=0,
			          DZ=0,),),
                     PRES_REP=(_F(GROUP_MA=('LIFT',),     # applying pressure on the top of the column
			          PRES=0.01,),),);

cine3=AFFE_CHAR_MECA(MODELE=FEMNolin,
                        PRES_REP=(_F(GROUP_MA=('BEND',),     # apply pressure on the side of the column
				     PRES=0.01,),),);

decal=1.;
					   
Rampe1=DEFI_FONCTION(NOM_PARA='INST',
                    VALE=(0.0 ,0.0 ,
                          1.0 ,1.0 ,
                          2.0 ,1.0 ,),
                    PROL_DROITE='CONSTANT',
                    PROL_GAUCHE='CONSTANT',);

Instant1=DEFI_LIST_REEL(DEBUT=0.0,
                        INTERVALLE=_F(JUSQU_A=decal,
                                      NOMBRE=20,),);

FRampe1=CALC_FONC_INTERP(FONCTION=Rampe1,
		         LIST_PARA=Instant1);
		  
Rampe2=DEFI_FONCTION(NOM_PARA='INST',
                    VALE=(decal+0.0 ,0.0 ,
                          decal+1.0 ,1.0 ,),
                    PROL_DROITE='CONSTANT',
                    PROL_GAUCHE='CONSTANT',);
					  				  
Instant2=DEFI_LIST_REEL(DEBUT=decal,
                        INTERVALLE=_F(JUSQU_A=decal+1.0,
                                      NOMBRE=20,),);

FRampe2=CALC_FONC_INTERP(FONCTION=Rampe2,
			 LIST_PARA=Instant2,);

SolNoLin=STAT_NON_LINE(MODELE=FEMNolin,
		       CHAM_MATER=Mat,
                       EXCIT=_F(CHARGE=cine1,
                                FONC_MULT=Rampe1,),
.
.
                       INCREMENT=_F(LIST_INST=Instant1,
				    INST_FIN=1.0,),
.
.
									  
SolNoLin=STAT_NON_LINE(MODELE=FEMNolin,
		       reuse=SolNoLin,
		       ETAT_INIT=_F(EVOL_NOLI=SolNoLin,),
		       CHAM_MATER=Mat,
                       EXCIT=(_F(CHARGE=cine3,
				 FONC_MULT=Rampe2,
				 TYPE_CHARGE='SUIV',),
			      _F(CHARGE=cine2,
				 FONC_MULT=Rampe2,
				 TYPE_CHARGE='DIDI',),),
.
.
                       INCREMENT=_F(LIST_INST=Instant2,
				    INST_FIN=2.0,),
.
.

Can one please suggest where I did wrong in the use of the EXCIT?

Thank you in advance for your support.

Regards,
Bhattarai

Last edited by bhattarai (2018-10-02 14:22:56)


Attachments:
lift.pdf, Size: 13.08 KiB, Downloads: 42

Offline

#2 2018-10-06 10:25:01

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 118

Re: Chained calculations with two STAT_NON_LINEs

Hello Bhattarai,
Do you want to keep the deformed shape of the first simulation without the stresses for the 2nd simulation?

K.


失敗は成功のもと (L'échec est la base de la réussite)

Offline

#3 2018-10-06 12:34:51

bhattarai
Member
Registered: 2014-01-31
Posts: 124

Re: Chained calculations with two STAT_NON_LINEs

Hi K,

Yes, I need to save the deformed shape and continue second deformation calculation after first one is over. Reference configuration for the second is the current configuration from the first calculation.

Thank you.

Bhattarai

Offline

#4 2018-10-06 16:30:49

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 118

Re: Chained calculations with two STAT_NON_LINEs

Hello,
I attached an example. The initial mesh is deformed with the displacements of the first analysis and the second analysis uses the deformed mesh.

Konyaro


Attachments:
poutre2steps.zip, Size: 35.16 KiB, Downloads: 33

失敗は成功のもと (L'échec est la base de la réussite)

Offline

#5 2018-10-08 11:24:55

bhattarai
Member
Registered: 2014-01-31
Posts: 124

Re: Chained calculations with two STAT_NON_LINEs

HI Konyaro,

Thank you for your effort. Unfortunately, I found certain difficulties with your files.

Mesh: It cannot be imported and generated error with some missing data (Salome version 7.6).

Command file: You have been using only one DEFI_LIST_INST for both STAT_NON_LINE.

I am not able to run the simulation because of the mesh file and I do not know what exactly is going on with the input comm and mesh file.

Is it possible to attach importable mesh file?

Thank you again.

Bhatarai

Offline

#6 2018-10-08 21:11:46

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 118

Re: Chained calculations with two STAT_NON_LINEs

Hello,
I uploaded new files:
- Mesh file saved in MED 3.0, I hope you can read it.
- Comm file corrected: the first IMPR_RESU which prints the result of the first analysis was after the second analysis...

Good luck,

Konyaro


Attachments:
poutre2steps_II.zip, Size: 38.29 KiB, Downloads: 28

失敗は成功のもと (L'échec est la base de la réussite)

Offline

#7 2018-10-10 15:27:32

bhattarai
Member
Registered: 2014-01-31
Posts: 124

Re: Chained calculations with two STAT_NON_LINEs

Hi Konyaro,

Just a quick question, did you encountered such an error during your calculation?

ERREUR A L'INTERPRETATION DANS ACCAS - INTERRUPTION
>> JDC.py : DEBUT RAPPORT
CR phase d'initialisation
   !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
   ! <S> Exception utilisateur levee mais pas interceptee.                  !
   ! Les bases sont fermees.                                                !
   ! Type de l'exception : error                                            !
   !                                                                        !
   !      Objet JEVEUX inexistant dans les bases ouvertes : >&&OP0070.DISC.     !
   ! .LINF<                                                                 !
   !      l'objet n'a pas été créé ou il a été détruit                      !
   !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
fin CR phase d'initialisation

Thank you.
Bhattarai


Attachments:
TEST.mess, Size: 38.91 KiB, Downloads: 21

Offline

#8 2018-10-11 07:09:18

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 118

Re: Chained calculations with two STAT_NON_LINEs

No, the oldest version of CA I have is 13.2 and it even works with that version... you should update to a new version.


失敗は成功のもと (L'échec est la base de la réussite)

Offline

#9 2018-10-11 13:20:23

bhattarai
Member
Registered: 2014-01-31
Posts: 124

Re: Chained calculations with two STAT_NON_LINEs

Thank you konyaro, the problem was my older version and it works now on v 13+. However, there might have been certain misunderstanding in my conversation. After testing your input files the simulation is successful though this was not what I needed. In the second AFFE_CHAR_MECA

cine02 = AFFE_CHAR_CINE(MECA_IMPO=(_F(DX=-1.0,
                                      GROUP_MA=('dispZ', )),
                                   _F(DX=0.0,
                                      DY=0.0,
                                      DZ=0.0,
                                      GROUP_MA=('fix01', ))),
                        MODELE=model1)

whatever direction of the prescribed displacement (DX, DY or DZ for GROUP_MA=('dispZ',)) I choose, the result is the same. It seems the second STAT_NON_LINE is never activated.

Also, I realize in your first reply:

Do you want to keep the deformed shape of the first simulation without the stresses for the 2nd simulation?

I needed to preserve the result of the first simulation and continue it as reference configuration for the second one. i.e., after a displacement of DZ=-1.0, in the first STAT_NON_LINE by,

cine01 = AFFE_CHAR_CINE(MECA_IMPO=(_F(DZ=-1.0,
                                      GROUP_MA=('dispZ', )),
.
.
my geometry should displace by -1.0 in Y-direction
cine01 = AFFE_CHAR_CINE(MECA_IMPO=(_F(DY=-1.0,
                                      GROUP_MA=('dispZ', )),
.
.
with second STAT_NON_LINE. It seems, with the provided input file, the result as described can not be achieved. What do you think?

Offline

#10 2018-10-12 21:11:36

konyaro
Member
From: Genève
Registered: 2016-04-02
Posts: 118

Re: Chained calculations with two STAT_NON_LINEs

Hello bhattarai,
The animation below shows the result of the files I attached before. I just increased the values of the displacements. The second analysis uses the deformed geometry of the last step of the first analysis. This is what I understood from your sketch.

Deformed shapes

Are you using AsterStudy? If yes, you must split the analysis into 2 stages because AsterStudy reorders the commands so the first IMPR_RESU will be moved to the end and will print the results on the deformed mesh.

Konyaro

Last edited by konyaro (2018-10-12 21:11:59)


Attachments:
2steps.gif, Size: 206.73 KiB, Downloads: 87

失敗は成功のもと (L'échec est la base de la réussite)

Offline

#11 2018-10-15 10:51:15

bhattarai
Member
Registered: 2014-01-31
Posts: 124

Re: Chained calculations with two STAT_NON_LINEs

Hello konyaro,

Yes, your two results looks reasonable if we change the direction of the prescribed displacement (DY or DZ for GROUP_MA=('dispZ',)). I am wondering why I could not get the same results with the same input file.

Nevertheless, I have attached the result in ogv format with different boundary conditions. Here, I have lifted the upper end of the beam vertically in first simulation. Afterwards, the lifted face is fixed and a horizontal pressure is applied on the beam laterally until the end of the second simulation. I could achieve this using commands, TYPE_CHARGE='SUIV' and TYPE_CHARGE='DIDI'.

At this moment, what I intended to achieve in the second simulation was not to fix the lifted face, instead apply a vertical pressure. The horizontal set of pressure on the lateral side of the beam is applied as it is.

Bhattarai


Attachments:
TEST.ogv, Size: 49.41 KiB, Downloads: 19

Offline

Board footer