Natural Convection ! Boussinesq Vs Ideal Gas density
 

Hi Everyone!

I am simulating cooling of a component by natural convection. The component is placed in rectangular box at middle location with rectangular big box inlet at bottom and outlet at top.

What should be density selections in create/edit material tab (ideal gas or Boussinesq)? Why operating density should be zero in case of ideal gas?

I run simulation for above two types of air density keeping gravity y=-9.81m/s (inlet is at y=0, outlet is at y>0)
For density as ideal gas , flow of air is downward (towards inlet)
For density as Boussinesq, flow of air is upward (towards outlet)
Avg velocity at inlet or outlet for ideal gas density is 1.5m/s and that of Boussinesq density its 0.3m/s (which seems okay for natural convection)

Can anyone please explain why there is difference in direction of velocity flow by just varying density of air in material tab?

Thank you in advance.

Hope someone take few minutes to reply ;):D

As of now I understood that density of fluid should NOT be selected as "Ideal Gas" In 'create/material' tab of Fluent for an OPEN system. For a CLOSED system with no inlet or outlet it will work.

Ideal gas can work for an open system... You are just doing something wrong.


With Boussinesq you do not need to specify any operating density. With ideal gas, you need to specified an operating density. Pressures in Fluent are a little funny and some bookkeeping is needed. You specify not the actual static pressure but with the hydrostatic part removed. Some people like to set the operating density to 0 so that they can set the static pressure on the boundary (just like setting 0 Pa operating pressure lets you specify absolute pressure instead of a gauge pressure).



So my question is, what are your boundary conditions? Don't think they are the same when "all I did was change density."

Thank you so much for reply.

With Boussinesq i did not specified operating density. For ideal gas, I tried specifying zero and non-zero density as well.

My boundary conditions are 'pressure inlet=0 pa' 'pressure outlet= 0' remaining 4 side of big box are adiabatic and some heat source term to the component cell zone. operating pressure= default (101325 pa)
All these BC are same for both cases.

For ideal gas as density case,
Changing "type" under boundary conditions from 'pressure inlet' and 'pressure outlet' to 'Wall' gives me flow upward.

Hi,
Have you found any solution, because iam also facing the same issue. Iam also trying to model natural convection around a pcb exposed to ambient conditions. when i used ideal gas flow vectors are drooping downwards. I expected they should move upwards and leave from outlet.
my model consists of inlet on LHS and outlet on RHS with rest as walls.

I got flow as per physical intuition with boussinesq model. but in that i got temp change >75degrees. so i cant conclude them as final results.

Any help would be greatly appreciated.

Thanks in advance.