thermal analysis with radiation problem

Hi.

I've been trying to run a thermal analysis including radiation. Initially I've simply set up a cuboid block with a heat flux in at one end and radiation at the other (since I can easily analytically check that it's working correctly).

If I use the linear wizard and then edit to include radiation, I get an error as it doesn't compute any radiation so the temperature isn't constrained (I checked this by adding an exchange alongside the radiation - temperature is then independent of radiation parameters).

So I then used a .comm file from a quadratic assembly analysis to edit my thermal .comm file for non-linear thermal analysis, but it seems it's failing to calculate any iterations. I've attached the flasheur file. I've looked through the relevant documentation and tried a few changes but nothing has helped.

I guess this is the relevant part of the flasheur file:

[code:1] Temp=THER_NON_LINE(EXCIT=_F(CHARGE=Config),
PARM_THETA=0.56999999999999995,
SOLVEUR=_F(RENUM='METIS',
STOP_SINGULIER='OUI',
METHODE='MULT_FRONT',
NPREC=8,
SYME='NON'),
NEWTON=_F(ITER_LINE_MAXI=0,
REAC_ITER=0,
RESI_LINE_RELA=1.E-3),
COMP_THER_NL=_F(RELATION='THER_NL'),
CONVERGENCE=_F(ITER_GLOB_MAXI=10),
MODELE=FemQuad,
CHAM_MATER=Mat,
);

--- NOMBRE TOTAL DE NOEUDS : 53893 DONT :
0 NOEUDS "LAGRANGE"
--- NOMBRE TOTAL D'EQUATIONS : 53893
--- TAILLE DU PROFIL MORSE DE LA TRIANGULAIRE SUPERIEURE (FORMAT SCR): 732364
--- DONC LA TAILLE DE LA MATRICE EST:
--- EN SYMETRIQUE NNZ= 732364
--- EN NON SYMETRIQUE NNZ= 1410835

****************************************************************************
* INSTANT: 0.000E+00 *
*************************************************************************************
* ITERATION RESIDU RESIDU ITERATION COEFFICIENT ACTUALISATION *
* RELATIF ABSOLU RECH. LIN. RECH. LIN. MATRICE *
* RESI_GLOB_RELA RESI_GLOB_MAXI RHO TANGENTE *
*************************************************************************************

!

---

!
! <EXCEPTION> <ELEMENTS2_63> !
! !
! !
! ! comportement non trouve ! !
! !
! !
! !
! !
!

---

![/code:1]

I'd be grateful for pointers as to how this analysis should be defined, or if anyone can point me at a complete .comm example for thermal analysis with radiation then that would be fine too.

Thanks.
Joe


Ah! One more thing! Is it possible to produce results of the flux flowing from the model but separated into radiation, and heat exchange from the model?



Post edited by: Joe Dowsett, at: 2010/07/15 13:44

Post edited by: Joe Dowsett, at: 2010/07/15 13:53<br /><br />Post edited by: Joe Dowsett, at: 2010/07/15 13:54

Hi Joe

! comportement non trouve !
means that Your material definition does not include the constitutive law which You ask in
THER_NON_LINE(. ..COMP_THER_NL=_F(RELATION='THER_NL')
to be used. Can You post Your complete .mess-file?

Regards,
mit freundlichen Grüßen,

Johannes Ackva

______________________________________________________________________
Ingenieurbüro für Mechanik
Dr.-Ing. Johannes Ackva
Markgrafenstr. 21
D 91717 Wassertrüdingen

www.code-aster.de
Training &amp; Support for NASTRAN and CODE-ASTER

Hi.

Thanks, that got me started at least. Yep, I had to change my material definition to THER_NL from THER. The parameters for this needed to be functions whereas I'm (for now at least) happy with constant values so have defined these functions by

[code:1]L_FUNC=DEFI_FONCTION(NOM_PARA='TEMP',
NOM_RESU='LAMBDA',
VALE=(-200.0,180.0,10000.0,180.0,),);
[/code:1]

The analysis then proceeded further, but I got the error

[code:1]You make a calculation of stationary heat without resolution and without either resolution transient.
Tips
Find the time discretization by keyword INCREMENT[/code:1]

(courtesy of google). Was a little surprised as surely there should be a stable solution - constant heat flux in one end, temp dependent radiation out at the other. Anyway, I then dug a little further into the documents and have now defined the Increment part of THERM_NO_LINE.

The analysis now runs but stops with temperature errors at different time intervals - I expanded my function definitions to encompass these, but I'm now getting temperatures of -1800 - something isn't right!

Thanks for any help.

Hi Joe,
You should try with a better initial temperature estimation, I guess (am not shure) a higher initial temp is better.

Please confirm or let me know if there remain problems

mit freundlichen Grüßen,

Johannes Ackva

______________________________________________________________________
Ingenieurbüro für Mechanik
Dr.-Ing. Johannes Ackva
Markgrafenstr. 21
D 91717 Wassertrüdingen

www.code-aster.de
Training &amp; Support for NASTRAN and CODE-ASTER

No, that didn't help. I even set it at the temperature that would be required at the radiating face for the power radiated to be equal to my power input (in this case it did manage to make it to the next time increment!).

I also added an exchange condition on the radiating face to see if it was a problem with the radiating condition. Again this didn't help.

I've dug through code asters forums too but couldn't find anything relevant relating to radiation.



Cheers
Joe

I You sent me mesh + .comm I could have today a more thorough look onto it. If so, You find my e-mail address on my website.

mit freundlichen Grüßen,

Johannes Ackva

______________________________________________________________________
Ingenieurbüro für Mechanik
Dr.-Ing. Johannes Ackva
Markgrafenstr. 21
D 91717 Wassertrüdingen

www.code-aster.de
Training &amp; Support for NASTRAN and CODE-ASTER