Natural Convection of a Solar Chimney

[Jaloboch]

Hi there,

I am new in this software and would really appreciate some guidance.

I want to simulate the natural convection inside a solar chimney using the boundary conditions 'Heat Flux'.

What I did was to draw the solar chimney model, enclosed by glass, with zinc at the bottom end and PVC at the top end. At the two ends, there is a 10 cm circular hole for the air to go in and go out.

The boundary conditions that I applied was heat flux (800 W/m^2) at the four sides of the chimney, pressure of 0 at both inlet and outlet, and a temperature BC of 35C at the inlet, to indicate an intake of air at 35C.

I feel like I'm lacking of some BCs as my results show an unusually high temperature. What should I do?

By the way, I ticked flow, heat transfer and radiation to be on. There's also a gravity and variable air.

Picture attached shows the model figure.

Thanks.

[apolo_vanderberg]

Jaloboch

For Natural Convection a large portion of the heat loss can be due to radiation. Having a Heat Flux condition on the exterior will bring in a set amount of heat however, the only method for heat to escape will be to convect/conduct to the interior air space. Those 4 glass walls that have the Heat Flux, will also radiate heat to the external surroundings (that we dont have modeled).

From the screen shot here are a couple other suggestions

Nice job settings ADV5!
Being Nat. Conv we would run this as Laminar (unless size was large enough that the Rayleigh number would be pushing us to Turbulent - we would not run Subsonic Compressible).

Be careful with the meshing as we will require sufficient mesh to capture any and all flow features inside the chimney.

Watch out for boundary conditions placement - namely the inlet temperature. We would prefer this to be assigned to the fluid only (such that the surface is unique to the fluid and not sharing an edge with neighboring solids). This can be done by adding a small fluid extension for the fluid and assigning the conditions there

For the External radiation, you could try adding a Radiation BC as well that would allow heat to radiate away from the model without having to include the external domain.

[Jaloboch]

Dear Apolo,

Should the glass tranmission be 0?

[Jaloboch]

Apolo, thank you for your guidance.

After taking into consideration on the guidance given by Apolo,

I tried running the simulation.

But the velocity results I get seems very weird as attached.

Why is this happening, although all the BCs are equally placed at four sides?

[apolo_vanderberg]

There are methods of representing radiation through the glass, for the time being its ok to keep it as is

The velocity profile is a function of your mesh. you will need to refine the mesh to capture that more effectively.