Bending curves along a beam/Courbe moment fléchissant [SOLVED/RESOLU]

pe89100

Hi everybody,

I'am trying to improve a static analysis of a beam on rigid supports. I would like to print some of the results like bending torque, and transverse force along the beam. However I did not succeed, I tried with POST_RELEVE_T and IMPR_TABLE but I do not really understand how they work. It seems that the results are only available at nodes (probably they are calculated at node only).

I attached the comm file and msh. I'm using windows STA 9.1 but I can also use STA10.1 with CAELinux.

Regards,

Paul.

Bon il semble que la PJ ne passe pas (2 ko pourtant), je réessaye plus tard...

Attached file:
PigVert.rar, 1979kb

pe89100

Dear all,

Is it possible to get a few information about the option to get the "internal force". Is it the option "SIEF_ELGA_DEPL" or the "EFGE_ELNO_DEPL" that correspond to the internal force? I understand that one is corresponding to the Gauss element although it is not really clear for me what is it in reality.

Second thing is that I would on the MF torques the curves looks likes stairs whereas the transverse force are straight... I save a *png picture in the archive below.

If somebody can help that would be fine, thanks in advance!

Paul.

Attached file:
PigVert.rar, 29969kb

jeanpierreaubry

bonjour

you are on the good track

SIEF_ELGA_DEPL are the forces in the beam in the local axis

i believe the gauss point are at the end node for beams like these

EFGE_ELNO_DEPL are the forces at nodes in the global axis

in this case the 2 coordinates system are the same (due to the beam lying along the x global axis)

the curves you produce do not like odd to me (i did not check the value though)

with which tool did you made them?

however their titles look strange to me

it is better nowadays to export a.med file from gmsh, you can nme the group within gmsh

jean pierre aubry

pe89100

Thanks for your support jean pierre,

jeanpierreaubry wrote:
the curves you produce do not like odd to me (i did not check the value though)

with which tool did you made them?

however their titles look strange to me

it is better nowadays to export a.med file from gmsh, you can nme the group within gmsh

jean pierre aubry

I have used GMSH to plot them, just switch the view space from 3D to 2D space in the option of the post pro module of GMSH.

I have tried to export the mesh as MED format but as I did not name manually my groups in GMSH (I use the number generated directly from GMSH) the script crash cause it does not now the named group GM101,... and every other group that I refer to with the gmsh file format.

The other try is import as GMSH and export as MED, in this way there is only the displacement plot when I open the result rmed file with GMSH (no VY, VZ, MFY, ...). However when I print the results in a RESU file those components are calculated!

I will give a try with complete import/export as MED files to see what happen.

Regards,

Paul

Attached file:
PigVert.tar.gz, 12939kb

jeanpierreaubry

when post processing a .med file Gmsh can only display the first component of the CHAM, that is the modulus of the vector, which very often is not enough

i procced that way :

geo in gmsh

name groups in .geo file

mesh

export .med file

calculate in aster

creating :

a .resu file,

a .med file, for Salome

a .pos file for Gmsh (this only possible in astk)

and i post process one or the other depending of what i want, also of the mood!

more and more i do some bits of post processing with STANLEY

if you are interested i am preparing a typical .comm file with some comments for a tutorial

pe89100

jeanpierreaubry wrote:
calculate in aster

creating :

a .resu file,

a .med file, for Salome

a .pos file for Gmsh (this only possible in astk)

and i post process one or the other depending of what i want, also of the mood!

more and more i do some bits of post processing with STANLEY

if you are interested i am preparing a typical .comm file with some comments for a tutorial

of course I'm interested!

On my side I was looking for a solution to plot directly the curve from aster in a picture format but it seems to me a little bit difficult...

At the moment I'm trying with option CREA_TABLE...

One more question :

When post with SALOME does the MFY direction vector indicate something. To be honnest I do not see what could be the link between their direction and the values of MFY. There is a picture below.

Paul.

Attached file:
MFY.png, 9122kb

jeanpierreaubry

here is a small problem with :

.geo

commented .com

it is the beginnings of a tutorial i want to write about :

Gmsh Code_Aster Salome-Meca

beam and cables structures, linear and non linear

Attached file:
tut101215.tar.gz, 9278kb

pe89100

I can not open the archive, it seems to be corrupted.

tcdonley

Hi - no problem with the archive, it has just been "double packed." Here's how to open it:

Open a terminal window, and cd to the directory where the file is saved.

Type:

gunzip tut101215.tar.gz

Now you can extract the tar file per your usual method.

Thanks for working on this!

- Tim D.

pe89100

Thanks for the tip Tim, I'm quite a beginner with LINUX!

Now I got the files I get always copy error, I do not yet understand why.

jeanpierreaubry

pe89100 wrote:
Now I got the files I get always copy error, I do not yet understand why.

when doing what?

pe89100

Running the analysis. I even can not read the mess file.

Maybe the flasheur file can help!

Attached file:
port1.o5397, 22038kb

jeanpierreaubry

how are you trying to run it

Salome-Meca?

or ASTK?

if you start Salome-Meca and create a proper "study case" with "port1.med" as mesh and "port1.comm" as command file it should work

did you copy these files and create the problem in a directory in which you do have write permission??

does not look like it

pe89100

I was trying to solve it with ASTK, and you were right with Salome it works fine.

By the way I almost finish the script command file and I will added later here.

Tanks!

pe89100

My *.comm file is almost finished but I still miss something...

I would like to create a table with the force reaction at the node where I constrained the shaft line (3 nodes P1, P7, P10).

However I was thinking to use the good syntax, but when I open the Table result there is only one line and it looks like that it is the sum reaction of all node and not the reaction at each node. Can somebody tell me what it is all about?

Thanks!

Paul.

Attached file:
PigVertcomm.tar.gz, 1444kb

jeanpierreaubry

the archive is unreadable, corrupted

however there are in my example port1.comm the lines that will print le reaction at every node in the .resu file

pe89100

Dear jeanpierre,

I think you encountered the same problem has me a few days ago while reading your archive.

You can look to the message of tcdonley just a few message before this one. You have to use gunzip first.

Of course I read your example and I think I reproduce the part of the comm which I'm interested in and however it does not work...

jeanpierreaubry

pe89100 wrote:
I think you encountered the same problem has me a few days ago while reading your archive.

You can look to the message of tcdonley just a few message before this one. You have to use gunzip first..

well, i know that issue and did it as it should be done

http://www.code-aster.org/forum2/viewtopic.php?id=12294

may be retry with a new archive

pe89100 wrote:
Of course I read your example and I think I reproduce the part of the comm which I'm interested in and however it does not work...

what means it does not work?

pe89100

Maybe it will work with this archive, I do not know why but the gunzip one was opening on my laptop.

In fact it is not really that it does not work but more than it does not produce what I want. In fact I would like to extract the reaction at each of my supporting node. With what I use it seems that I have the sum of reaction at all mentionned node (P1, P7, P10). This was my meaning.

Attached file:
PigVertcomm.tar, 10240kb

jeanpierreaubry

well, quite simple

instead of making one single table with GROUP_NO=('P1','P7','P10',),

you have to make three different tables with GROUP_NO=('P1',),...GROUP_NO=('P7',),...GROUP_NO=('P10',),

if you do not print in a table but in direct FORMAT='RESULTAT', you can use GROUP_NO=('P1','P7','P10',),

and you vill have a much nicer printout to my opinion with the node numbers