How is surface temperature computed ?

[burkhart68]

In dealing with surface pool evaporation, I need to understand how is the surface temperature computed by CS.
When I force it through a user formula Ts(t) in the IHM, it seems to be recomputed and doesn't correspond to the imposed one.
When I try to force a Dirichlet value for the variable itempb (ground temperature) in a rough Wall boundary condition, CS refuses to run.
Thanks for clarifying this point !
Stephane Burkhart

[YvanFournier]

Hello,

Are you trying to force a Dirichlet value with the GUI ? Or using another (mixed) BC type ? If you do this and activate the visualization of surface temperature, the result should be the same. Otherwise, this might be a bug (ensuring the values are consistent is tricky, and we have had issues in the past due to a mismatch in which parts of the time step things are updated, so if this type of bug has resurfaced, wee need to check this. Otherwise, with Neumann or mixed BC types, T* and T+ values are computed when converting the prescribed BC values to coefa/b values (in cs_boundary_conditions_set_coeffs_turb for v9.0+, clpturb for older versions).
This is then used in postprocessing utility functions (in cs_function_default.cpp) to reconstruct the boundary temperature.

Also, do you have a wall law for this surface ? It is possible an issue would occur only with some models...

[burkhart68]

Merci Yvan, je vais regarder ça de plus près. En fait je cherchais à forcer une valeur de température variable dans le temps sur une condition de rough wall, qui attendait plutôt un flux de chaleur visiblement. J’ai essayé 1) par formule dans l’IHM (galère avec le formatage et ne fait pas ce que je lui dis quand je visualise la température au sol dans RESULTS_BOUNDARY.case), et 2) par routine utilisateur (plante en forçant la variable itempb). J’utilise la v6.0, en keps linear avec loi de parois rugueuse (2-scale model + Rough wall) Je vais regarder du côté des conditions mixtes… Je te tiendrai informé. Stéphane De : Yvan Fournier ***@***.***> Envoyé : jeudi 28 août 2025 12:21 À : code-saturne/code_saturne ***@***.***> Cc : BURKHART Stephane ICT ***@***.***>; Author ***@***.***> Objet : Re: [code-saturne/code_saturne] How is surface temperature computed ? (Issue #151) La plateforme SISMEL a supprimé tout ou une partie de ce message en raison de la présence de contenu actif potentiellement dangereux. [Image supprimée par l'expéditeur.]YvanFournier left a comment (code-saturne/code_saturne#151)<#151 (comment)> Hello, Are you trying to force a Dirichlet value with the GUI ? Or using another (mixed) BC type ? If you do this and activate the visualization of surface temperature, the result should be the same. Otherwise, this might be a bug (ensuring the values are consistent is tricky, and we have had issues in the past due to a mismatch in which parts of the time step things are updated, so if this type of bug has resurfaced, wee need to check this. Otherwise, with Neumann or mixed BC types, T* and T+ values are computed when converting the prescribed BC values to coefa/b values (in cs_boundary_conditions_set_coeffs_turb for v9.0+, clpturb for older versions). This is then used in postprocessing utility functions (in cs_function_default.cpp) to reconstruct the boundary temperature. Also, do you have a wall law for this surface ? It is possible an issue would occur only with some models... — Reply to this email directly, view it on GitHub<#151 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AOLGIBU4QKEGRG3FTANZH3T3P3JXRAVCNFSM6AAAAACFAONZY2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTEMZSHA4TGOJRGM>. You are receiving this because you authored the thread.Message ID: ***@***.***>