Coding Boussinesq approximation

[Ruonan]

Dear developers,

Please could you help me with this question?

I did the "heated square cavity flow" tutorial case recently, but I am confused about the way it applies the Boussinesq approximation.

I suppose the Boussinesq approximation only changes density in the gravity term, from a constant to a temperature-dependent variable. But what I did is change the density to a variable in the subroutine, so is it correct that the density became a variable not only in the gravity term, but also in other terms in the momentum equation? Could you please help me to confirm this?

This problem was also discussed in this post but it seems no obvious conclusion was made.
viewtopic.php?f=2&t=1570&p=8549&hilit=c ... nesq#p8549

Many thanks,
Ruonan

[Yvan Fournier]

Hello,

I will be happy of other developers more familiar with this part of the code confirm this, but yes, from recent discussions we have had about this subject, is is possible to use the Boussinesq approximation for buoyancy terms but also have a slightly varying density for the other terms in the momentum equation.

I do not know whether in this case there is an advantage to using Boussinesq (idilat = 0), over the default dilatable algorithm (idialt = 1).

Best regards,

Yvan