code_saturne User's Forum

Hello,

There appears to be a typo error in 'condli.f90' file (CS V4.2.1). The term in bold.

------------------------------------------------------------------------------------------------
condli.f90
------------------------------------------------------------------------------------------------
! The implicit term is added after having updated the velocity
forbr(1,ifac) = ( cofafu(1,ifac) &
+ cofbfu(1,1,ifac) * velipb(ifac,1) &
+ cofbfu(1,2,ifac) * velipb(ifac,2) &
+ cofbfu(1,3,ifac) * velipb(ifac,3) )*srfbnf
forbr(2,ifac) = ( cofafu(2,ifac) &
+ cofbfu(2,1,ifac) * velipb(ifac,1) &
+ cofbfu(2,2,ifac) * velipb(ifac,2) &
+ cofbfu(2,3,ifac) * velipb(ifac,2))*srfbnf
forbr(3,ifac) = ( coefau(3,ifac) &
+ cofbfu(3,1,ifac) * velipb(ifac,1) &
+ cofbfu(3,2,ifac) * velipb(ifac,2) &
+ cofbfu(3,3,ifac) * velipb(ifac,3) )*srfbnf
------------------------------------------------------------------------------------------------

Thanks,
Vilas

Hello Vilas,

Yes, this actually impacts all version since 3.0, and was coincidently discovered/fixed very recently (following another thread in this forum) For version 4.2, the fix is in the "subversion" branch, so 4.2.2 will contain it, but we need to build a bug-fix release to make it more easily available...

With the bug, the z-term of post processed boundary efforts/stresses is completely off...

Thanks for the feedback in any case.

Best regards,

Yvan