#1 2022-01-07 14:44:27

Romans1988

Member

Registered: 2021-11-08

Posts: 11

[SOLVED] - Echange and Maillage

Hi,
i 'am trying to perform a static thermal analisys in a single 3D body (no assembly) with a T imposed in two differents areas and surface Echange Boundary Condition in another.
I am obtaining the following error:
   !----------------------------------------------------------------------!
   ! <EXCEPTION> <MODELISA6_96>                                           !
   !                                                                      !
   !  les 12478 mailles imprimées ci-dessus n'appartiennent pas au modèle !
   !  et pourtant elles ont été affectées dans le mot-clé facteur :       !
   ! ECHANGE                                                              !
   !----------------------------------------------------------------------!

" les 12478 mailles" refering to the Echange BC.

Removing the echange boundary the calculation go to result.

I don't understand where is the error. I know that it is in echange BC.
In attachment the data file.

Thanks to al for the answer.

Last edited by Romans1988 (2022-01-11 17:54:48)

---

Attachments:
Thermal_Posted_2022_01_07_1433.rar, Size: 1.35 MiB, Downloads: 181

mf

Member

Registered: 2019-06-18

Posts: 407

Re: [SOLVED] - Echange and Maillage

Hello,

the error is in AFFE_MODELE, your face group 'Convection' is not part of the 3D group 'Pump Housing', thus it was not part of the model.

So changing AFFE_MODELE to:

model = AFFE_MODELE(
  AFFE=_F(
    MODELISATION='3D',
    PHENOMENE='THERMIQUE',
    TOUT='OUI'
  ),
  MAILLAGE=mesh
)

will do the trick,

Mario.

Last edited by mf (2022-01-07 15:31:30)

Romans1988

Member

Registered: 2021-11-08

Posts: 11

Re: [SOLVED] - Echange and Maillage

Hi Mario,
it's works. But i don't understand how. The Convection Area is a group of element faces of the 3D model.
Also if i impose a termal load (with ddl_temp) to the convection zone, i obtain the results without any error.
Instead if i use echange BC also to another area (for example in the zone calls T1 or T2) that surface automatically is cut out from the model.
I think that there is something that i don't have understand in the echange BC.
Thanks for the reply.

Last edited by Romans1988 (2022-01-08 10:25:46)

AsterO'dactyle

Administrator

Registered: 2007-11-29

Posts: 464

Re: [SOLVED] - Echange and Maillage

Hello,

Back to the finite element theory !

See  R3.06.02.

ECHANGE is solved by weak approach (integral), you need finite element on skin to make numerical integration

Dirichlet condition (temperature) DDL_TEMP is imposed by strong method, at nodes, without integral, you don't need skin elements

---

Code_Asterの開発者