×

Notice

The forum is in read only mode.

problem solving a shell

  • Borja AP
  • Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
15 years 3 months ago #3834 by Borja AP
problem solving a shell was created by Borja AP
Hello everyone

I'am a begginer in salome, I'am trying to solve an simple plate under pressure in one side, with pinned borders, and as I am modelling 1/4th of the plate, the DOF of the two sides remaining, are restricted in order to behave symetrically.

What I cannot fix is the "MODELISA7_77" exception, I have taken a look to similar posts but I do not figure out how to solve it.

If any of you could help me on this I would really appreciate it.

Thank you very much in advance.

PS: You will find the .comm .med .hdf and error files attached.

Attachment fixed_plate.zip not found

Attachments:
More
15 years 3 months ago #3837 by kwou
Replied by kwou on topic Re:problem solving a shell
hoi Borja

the problem is that your ''borde'' is a group of nodes and not a part of a geometrical entity. So you have to change GROUP_MA='borde' to GROUP_NO='borde'. The same is valid for sym_x and sym_y. ''presion'' is probably alright. See code below.

[code:1]
malla=DEFI_GROUP(reuse = malla,
MAILLAGE=malla,
ALARME='OUI',
CREA_GROUP_MA=(_F(NOM='borde',
GROUP_NO='borde'), #<--
_F(NOM='sim_y',
GROUP_NO='sim_y'), #<--
_F(NOM='sim_x',
GROUP_NO='sim_x'), #<--
_F(NOM='presion',
GROUP_MA='presion')),
);

#from message file:
#le GROUP_MA borde ne fait pas partie du maillage: malla
[/code:1]

Maybe you also have to change the CREA_GROUP_MA --> CREA_GROUP_NO. See the message file to be sure.
In the code below some alternatives I used in one of my command files, are given

[code:1]
MeshQuad=DEFI_GROUP(reuse =malla,
MAILLAGE=malla,
#[CREA_GROUP_MA=_F(NOM='nborde',GROUP_MA=('borde',),),]
#[CREA_GROUP_NO=_F(GROUP_MA='borde',),]
CREA_GROUP_NO=_F(NOM='nborde',GROUP_MA='borde',),);
[/code:1]

Have fun - kind regards - kees<br /><br />Post edited by: Kees Wouters, at: 2010/02/07 22:00

Interest: structural mechanics, solar energy (picture at 'my location' shows too little pv panels)

--
kind regards - kees
  • Borja AP
  • Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
15 years 3 months ago #3841 by Borja AP
Replied by Borja AP on topic Re:problem solving a shell
Hello Kees,

First of all, thank you for you comments and help.

I have been trying some changes, and I end up with a .mess file which looks better, the alarm message says that may be, some nodes doesn't know how to move on dz direction, I guess, I have then, two questions, could I find somewhere documentation about the finite elements used by salome and their DOF per node?? and, it is because of this i'm not getting good results?

As always, I attached here the files generated by the analysis.

and my last question, do I have to go through ASTK always? or meshing in SALOME and editing with EFICAS is enough to do any kind of analysis?

Attachment fixed_plate2.zip not found

Attachments:
More
15 years 3 months ago #3843 by kwou
Replied by kwou on topic Re:problem solving a shell
Hoi Borja

I think you have to change your model from 'AXIS' to 'DKT' or 'COQUE_3D' to model a shell. I havenot used 'AXIS' before, but from the documentation I understand that this is used to expand a 2D model around an axis to get a 3D model. So in 'AXIS' only x (radial) and y (axial?) displacements are valid: z displacement does not exist. Hence the error in your mess file.

MODELO=AFFE_MODELE(MAILLAGE=malla,AFFE=_F(TOUT='OUI',PHENOMENE='MECANIQUE',MODELISATION='AXIS',),);

so change MODELISATION='AXIS' to MODELISATION='DKT'/'COQUE_3D' to model shell elements. See e.g. www.caelinux.org/wiki/index.php/Contrib:...outers/platedynamics

So the elements generated in Salome need to be ''modeled'' in CA. In Salome all nodes have just 3D coordinates (x,y,z). And elements in Salome can be either linear or quadratic, you can easily change between those two shapes in it. However in the CA-commmand file the modelisation and the choice of the element type takes place. Eg a linear triangle can be modeled into a 'DKT' shell element that has 6 degree of freedoms (dof, DDL): 3 displacements and 3 rotations. You have to look into the CA documentation for all the element types and their characteristics. At first this is the most tricky part I found. I paste a command string into the ''documentation'' search path and start reading and follow links. For eg paste AFFE_MODELE and the first hit will be: www.code-aster.org/V2/doc/v9/man_u/u4/u4.41.01.pdf .

If you have further questions, or problems, let us know. We try to solve them.

kind regards - kees<br /><br />Post edited by: Kees Wouters, at: 2010/02/08 23:53

Interest: structural mechanics, solar energy (picture at 'my location' shows too little pv panels)

--
kind regards - kees
  • Borja AP
  • Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
15 years 3 months ago #3845 by Borja AP
Replied by Borja AP on topic Re:problem solving a shell
HELLO kEES

Ok, in further analysis attempts, I promise I will take a look at the documentation, it is that I'am quite exited of finding this fantastic open_source CAE software, difficult to find in the linux world!

Thank you again, and congratullations for all what you and people of caelinux are doing!

Salut!
  • Borja AP
  • Topic Author
  • Offline
  • Junior Member
  • Junior Member
More
15 years 2 months ago #3885 by Borja AP
Replied by Borja AP on topic Re:problem solving a shell
Kees Wouters wrote:

Hoi Borja

I think you have to change your model from 'AXIS' to 'DKT' or 'COQUE_3D' to model a shell. I havenot used 'AXIS' before, but from the documentation I understand that this is used to expand a 2D model around an axis to get a 3D model. So in 'AXIS' only x (radial) and y (axial?) displacements are valid: z displacement does not exist. Hence the error in your mess file.

MODELO=AFFE_MODELE(MAILLAGE=malla,AFFE=_F(TOUT='OUI',PHENOMENE='MECANIQUE',MODELISATION='AXIS',),);

so change MODELISATION='AXIS' to MODELISATION='DKT'/'COQUE_3D' to model shell elements. See e.g. www.caelinux.org/wiki/index.php/Contrib:...outers/platedynamics

So the elements generated in Salome need to be ''modeled'' in CA. In Salome all nodes have just 3D coordinates (x,y,z). And elements in Salome can be either linear or quadratic, you can easily change between those two shapes in it. However in the CA-commmand file the modelisation and the choice of the element type takes place. Eg a linear triangle can be modeled into a 'DKT' shell element that has 6 degree of freedoms (dof, DDL): 3 displacements and 3 rotations. You have to look into the CA documentation for all the element types and their characteristics. At first this is the most tricky part I found. I paste a command string into the ''documentation'' search path and start reading and follow links. For eg paste AFFE_MODELE and the first hit will be: www.code-aster.org/V2/doc/v9/man_u/u4/u4.41.01.pdf .

If you have further questions, or problems, let us know. We try to solve them.

kind regards - kees&lt;br /&gt;&lt;br /&gt;Post edited by: Kees Wouters, at: 2010/02/08 23:53


Hello again Kees!

I've been trying to solve your shell dynamic tutorial from www.caelinux.org/wiki/index.php/Contrib:...outers/platedynamics but even though I've used your code I get the COPY_ERROR message, I've been looking for a solution but I cannot find it.

I would really appreciate if you could take a look on the files I have herein attached.

Thanks in advance.

Bap

Attachment shelldynamics.zip not found

Attachments:
Moderators: catux
Time to create page: 0.145 seconds
Powered by Kunena Forum