CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Heat conduction in a solid domain (http://www.cfd-online.com/Forums/cfx/19900-heat-conduction-solid-domain.html)

S. Balasubramanyam September 18, 2003 06:22

Heat conduction in a solid domain
 
Hi, I am trying to solve a simple heat conduction problem using CFX5.6. I am getting an error message which says that there is no fluid domain. I have just taken a plate which has one hot face and 5 adiabatic faces with lower initial temperature. There is no fluid domain in the problem. Is it necessary that we must have a fluid domain in each problem? I request anyone who used CFX, to help me in understanding, how to handle CFX5.6 to solve this problem. Thanks in advance.

Robin September 18, 2003 11:01

Re: Heat conduction in a solid domain
 
Yes, just setup a fluid domain as normal but:

1. Set your velocity initial guess to zero, 2. Set your advection scheme to upwind to avoid vertex gradient calculations, 3. Set the expert parameter "solve fluids = f" to turn of the hydro equations, 4. Set the turbulence model to laminar.

You can also do all this and include CHT solids if you have more than one solid domain.

Regards, Robin

Märta-Karin September 26, 2003 04:52

Re: Heat conduction in a solid domain
 
Hi Robin

I am trying to model the cooling of a warm ring by a gas flow and I would like to get the temperature field inside the ring. But when I run the transient simulation the temperature field inside the ring is constant which it should not be.

Do you have an idea of what is wrong ?

Thanks a lot.

Märta

Robin September 26, 2003 16:12

Re: Heat conduction in a solid domain
 
Hi Märta,

Transient CHT simulations are really quite impractical because the timescale of the fluid dynamics is significantly shorter than that of the heat transfer through the solid. The temperature in your solid probably isn't constant, it is just changing so slowly that you are not seeing it.

If you don't expect the flowfield to change much, you could freeze the momentum equations (set expert parameter 'solve fluids = f' and use a much larger timestep. This would then solve the heat transfer and transport in the fluid, but the flowfield would remain fixed. If you do expect the flowfield to change significantly, I recommend finding yourself a good novel, as you may have to wait a while :)

Note that in a steady state simulation you would typically use a timestep 10x or 100x larger for your solid than your fluid, and you are already using a significantly larger timestep than required for a transient simulation.

Regards, Robin


Märta-Karin September 29, 2003 03:49

Re: Heat conduction in a solid domain
 
Hi Robin!

Thanks for your answer, it was very helpful and quite correct about the temperature changes.

I have another idea about how to make the solution go faster - I want to calculate the solution in steady state and then interpolate the solution on the gtm file I use for the transient runs. Though, this means that the mesh in the steady state and the transient have to be exactly the same, and I can't find out how to set the boundaries on the ring in the steady state case in order to keep the ring temperature constant.

Do you know how to set the ring to one temperature? It seems to me as it is not possible to use domain interfaces in this case.

Best Regards, Märta-Karin

Amit Katiyar October 10, 2003 13:52

Re: Heat conduction in a solid domain
 
Hi Robin,

I am also solving a similar problem S. Balasubramanyam. I followed your suggestions but i am having error as:<< There are domain location parameters that have been used more than once in this problem.>> This is because i have only one domain in my geometry.

So please guide in this context.

thanks amit

Robin October 10, 2003 16:03

Re: Heat conduction in a solid domain
 
Hi Amit,

You are probably referencing the same region for two boundary conditions.

Robin

Amit Katiyar October 10, 2003 17:34

Re: Heat conduction in a solid domain
 
hi Robin,

Thanks for replying. Actually I will have to refer same domain for both because I have only one 3-D domain in my geometry. My geometry is solid cube having bottom at 500 K and all other surfaces at 300 K and I want to see heat transfered by conduction.

So Guide me. Amit

Amit Katiyar October 11, 2003 13:36

Re: Heat conduction in a solid domain
 
Dear experts,

Please tell me the difference between static temperature and total temperature. I found them while specifying opening boundary conditions.

Thanks.

Amit

Amit Katiyar October 12, 2003 05:11

guide me
 
Dear experts,

Please tell me the difference between static temperature and total temperature. I found them while specifying opening boundary conditions.

Thanks.

Amit

Robin October 14, 2003 08:57

Re: guide me
 
Hi Amit,

The static temperature is that which you would get if you measured the temperature of the moving fluid. The total temperature is the temperature you would measure if the fluid was decelerated isentropically to zero velocity (relative to the reference frame), such as at a stagnation point (sometimes referred to as the stagnation temperature).

Regards, Robin


All times are GMT -4. The time now is 22:58.