[Dino9111]

Hi together,

I would like to simulate just a area (box with dx=50m x dy=50m x dz=25m) with compressible air using opening boundary at the sides and a wall at the bottom. Intention is to create a realistic atmosphere with a negative temperature gradient with increasing height without any flow (or almost).

Initial and boundary condition for temperature: T(y)=288K-(0.976 K/100m)*y
(using adiabatic temperature gradient of atmosphere g_ad)
ref. pressure: 1 atm at (0,0,0)
If I set "relative pressure" at "opening pressure and drn." at 0 Pa, I always get a flow downwards at the center of the box (when compressibility and byouyancy is enabled - I need both effects in later simulations).
I'm not sure but as far as I understood the doc, I should define hydrostatic pressure p(y) at my opening boundaries.
--> Initial and boundary condition for pressure: p(y)-pref with
p(y)=p0*(1-(g_ad*y/T0)^(M_air*g/(R*g_ad)) (barometric formula; p0,T0 at y=0)

But then, in the center there is a flow upwards.

Has anyone an idea how to set boundary conditions etc. correctly?
Thanks a lot in advance!
pref_zero_Temp.jpg

pref_zero_velocity.jpg

pref_baromform_Temp.jpg

[evcelica]

Translational periodic conditions might be suitable here.

[Dino9111]

Hi evcelica,

first of all thanks - it works qute well. But unfortunately then I can't define a wind profile at the sides (for later siimulations).
Isn't there any possibility to define relative pressure at an opening correctly (so that there is no flow initiated by pressure differences)?
And why doesn't relative pressure = 0 not work? No pressure difference-->no flow?!

[Dino9111]

OK i finally solved it...
for those who are interested:
You have to define for the realtive pressure at the opening boundary: p(y) - pref + density*g*y because p(y) is the absolute pressure, but the pressure you have to define at the opening boundary is interpreted to be without the hydrostatic part.