Subdomain or Interfaces (CHT)
 

Dear CFX users,


I am simulating (SST) a multi domain problem (CHT) described as:

i have a solid domain say rectangular channel of some thickness (solar absorber) and i have a fluid domain (air) flowing through it and i have interfaces between solid and fluid.

The solid gets heated upto 1000c through solar radiation (incoming heat flux) and the heat is transferd from solid to fluid. I have included heat transfer model and radiation model for the solid and only heat transfer model for fluid (no radiation).

one way i applied is boundary conditions (heat flux) on solid interfaces because i dont have wall boundary option on solid, as SOUCES----> RADIATION SOURCE----->HEAT FLUX W/m2 (top and inside channel as decreasing flux in the form of expression.) and i simulated but i didnt get good domain imbalances for solid say 50% and also the temperatures are too high say 2500c and it is going on increasing for time steps.

2nd way is i have created a subdomain for the solid and i applied boundary condition as SOURCES---->RADIATION SOURCE----->HEAT FLUX W/M3
(very high heat flux compared to that of interfaces as decreasing in the form of expression in the subdomain) and simulated, got good convergence and the temperatures are also in range as for reality.

My question is what is the difference between appling flux on interfaces (W/M2) and subdomain (W/M3) and which i have to apply. Why i am getting high temperatures on interfaces rather than in subdomain. But in reality the solar radiation is falling on top surface and some portion of inside channel(interfaces in my model).

and one more question is how do the solver calculates the temperatures in both cases as heat flux bc (actually i searched in the user guide i didnt find) and last what is the difference between RADIATION SOURCE and ENERGY SOURCE in the soucres tab.

Bit lenghty sorry......

Thanks in advance.

Hi Sir,

:confused: I also will do this kind of simulation (fins mounted on the heated rectangular plate in laboratory with the presence of the air around, using the air stagnation condition, or else, heat transfer coefficient = 20 [W m^-2 K^-1] and outside temperature = 25 [C]).

I have some doubts concerning of this problem, if you can help me. Where did you do your grid and what kind of grid did you use (hexahedrical or tetrahedrical meshes)? If you use tetra mesh, how did you build these 2D interfaces between the solid and fluid domains? Did you use CFX-Mesh? Is it possible to impose a boundary condition , for example, "Heat Flux in" in a 2D interface? I say this because my heated plate is between an isolated material and the base of my fin.

I will thank you for the help,
Rogerio.

Hi,

For CHT simulations using residuals as the only convergence critereon is often inadequate. You need to converge on both residuals and imbalances. Your report that you have a 50% heat imbalance in one body confirms you have not adequately converged.

It does not matter what type of mesh you use for this type of analysis. If also does not matter whether they are linked by 1to1 or GGI boundaries. In version 12 the old issue of 1to1 vs GGI interfaces for CHT boundaries has been fixed so this is not a consideration any more.

Glenn Horrocks

Hi Rogerio,

I did by using cfx mesh (y+ <1) and you can build any type of interfaces between multiple domains just create a domain interface between two domains(fluid and solid) and set the 2d regions for the domain interfaces and thats it you will get automatically two interfaces in fluid and solid and apply the boundary conditions.


sandeep

Hi Glenn Horrocks,

Thanks for your reply, actually i found the solution.

sandeep

Hi,

What was the solution?

Glenn Horrocks

Thanks sandeep!
I will try to do that!
:cool:

Hi,

you said: "...just create a domain interface* between two domains(fluid and solid)...", but i have one question Sir. Is this a 3D domain interface*? I never heard about it, a 3D domain for an interface. I think that an interface doesn´t have a thickness. Could you please explain better for us? Thanks.

Rogerio.

boundary condition on the interfaces
 

For this kind of problem, what kind of boundary conditions to you put on these interfaces, in a CHT simulation using ANSYS CFX 12 software?

Rogerio.