setting of connection of shells

yasuo · 2023-11-18 13:32:40

Excuse me.

I am trying to simulate of shell model.
The model is subjected to a linear load in the X-axis direction from the left of upper shell and fixed at the right of under shell.

But it does not simulate anyway.
Perhaps there are problems with the settings of connection between the upper, lower and central shells, but I am not sure.

Would you like to some advice please?

Attachments:
1118.zip, Size: 23.73 KiB, Downloads: 22

jeanpierreaubry · 2023-11-18 15:37:00

hello

the error in the message file means that surface right is free floating in space
so does surface target

what are supposed to do the two instance of LIAISON_SOLIDE ?

and there are a few other things which i do not understand

jean pierre aubry

k_zurawski · 2023-11-19 12:51:33

I haven't seen your model yet, but I know at least two ways of connecting shells. Merge nodes and elements in SMESH module or LIAISON_DDL* between groups. LIAISON_SOLIDE I'd only use when I want to transfer distant load, moment applied to node or rigid connection between specific two groups, for example like RBE2 spider.

*I made a mistake, I meant LIAISON_MAIL.

Last edited by k_zurawski (2023-11-19 17:10:57)

k_zurawski · 2023-11-19 17:18:51

LIAISON_SOLIDE=(_F(GROUP_NO=('r', )),

_F(GROUP_NO=('l', ))),

I'd use LIAISON_SOLIDE as separate AFFE_CHAR_MECA and invoke it separately in EXCIT (but this is my preference only). And you have a mistake in my opinion. _F(GROUP_NO=('r', ) should indicate both "contact" pairs.

Rigid=AFFE_CHAR_MECA(MODELE=B,
LIAISON_SOLIDE=_F(GROUP_NO=('r','l',),),
)

Last edited by k_zurawski (2023-11-19 17:34:39)

k_zurawski · 2023-11-19 17:27:31

In case of glue contact between meshes:

Glued=AFFE_CHAR_MECA(MODELE=B,
LIAISON_MAIL=_F(GROUP_MA_MAIT='liner',
GROUP_MA_ESCL='tagr',
TYPE_RACCORD='MASSIF', (check in documentation what correct RACCORD for shell is)
),)

this is for one instance, you have to add second glued pair.

k_zurawski · 2023-11-19 17:57:14

Let me know if you need more assistance. We can go step by step with the command file.

jeanpierreaubry · 2023-11-20 14:48:26

hello

find attached is a modified .comm file that (i hope!) solves the problem

with a more logical mesh the problem would be much simpler

but please read the documentation for each command before using it to understand what i does
you may also read my book

jean pierre aubry

Last edited by jeanpierreaubry (2023-11-20 17:08:35)

Attachments:
set.comm, Size: 3.54 KiB, Downloads: 23

k_zurawski · 2023-11-20 16:13:00

One question, listr and func0 are defined in this case to divide totla load to 20 increments? Am I right?

jeanpierreaubry · 2023-11-20 16:23:18

k_zurawski wrote:

One question, listr and func0 are defined in this case to divide totla load to 20 increments? Am I right?

it looks like that
but i do not know what the writer of the command file intend to do with this feature as the 20 steps are purely linear (proportionnal)

k_zurawski · 2023-11-20 16:25:27

Maybe to check at which step the stress is near yield?

jeanpierreaubry · 2023-11-20 16:42:25

k_zurawski wrote:

Maybe to check at which step the stress is near yield?

maybe though with MECA_STATIQUE it is strictly proportionnal
and with E=16000.0, i guess the parts are in wood
construction code do not mention yield stress for wood!

k_zurawski · 2023-11-25 22:31:19

Jean Pierre,
You have chosen NIVE_COUCHE='MOY' in POST_CHAMP, why not SUP or INF?

jeanpierreaubry · 2023-11-26 10:08:11

Krzysztof,

just my choice
it shows the stress state not taking into account local bending
of course a real life study shoud include SUP an INF as well

as another point
i am annoyed that all these remarks do not call any answer from the original poster
who did not bother to introduce himsef by the way !

Last edited by jeanpierreaubry (2023-11-26 10:08:44)

k_zurawski · 2023-11-26 10:42:11

Correct me if I'm wrong, but in this case the results for all the three layers should be the same?

As for the second point.
I can't speak for others, but despite few fierce battles I'm loosing when it comes to patience. I think it's the sign of of our times that we want to have all the information right away on a silver plate.
If he got discouraged at least I benefited from this conversation. So please don't feel that you wasted your time.
Thank you for sharing all your experience here and with your publication.

jeanpierreaubry · 2023-11-26 15:34:44

k_zurawski wrote:

Correct me if I'm wrong, but in this case the results for all the three layers should be the same?

of course for this very (plane) probleme, i answered a bit too fast

Thank you for sharing all your experience here and with your publication.

thanks to you

#1 2023-11-18 13:32:40

setting of connection of shells

#2 2023-11-18 15:37:00

Re: setting of connection of shells

#3 2023-11-19 12:51:33

Re: setting of connection of shells

#4 2023-11-19 17:18:51

Re: setting of connection of shells

#5 2023-11-19 17:27:31

Re: setting of connection of shells

#6 2023-11-19 17:57:14

Re: setting of connection of shells

#7 2023-11-20 14:48:26

Re: setting of connection of shells

#8 2023-11-20 16:13:00

Re: setting of connection of shells

#9 2023-11-20 16:23:18

Re: setting of connection of shells

#10 2023-11-20 16:25:27

Re: setting of connection of shells

#11 2023-11-20 16:42:25

Re: setting of connection of shells

#12 2023-11-25 22:31:19

Re: setting of connection of shells

#13 2023-11-26 10:08:11

Re: setting of connection of shells

#14 2023-11-26 10:42:11

Re: setting of connection of shells

#15 2023-11-26 15:34:44

Re: setting of connection of shells

Board footer