[aja1345]

Hi,


Is it possible to apply a Fixed temperature to a Volume of air in CFX?


I am simulating a problem that air is around a material and this air(volume) have a Fixed temperature in this problem. In fact, the material temperature rise because of this hot air volume .


Thanks.

[Opaque]

You can make a domain isothermal, i.e., set the heat transfer model to isothermal

[evcelica]

Yes, that is called a Dirichlet boundary condition. But it doesn't sound like a realistic setup to me.

You would apply a volumetric heat source (subdomain) that is a function of local temperature and intended temperature. Cp * Density *(T.intended - T)/1[s] with a source term coefficient of -1[s].

Again, this does not sound realistic if you are intending to simulate heat transfer.

[aja1345]

Originally Posted by Opaque

You can make a domain isothermal, i.e., set the heat transfer model to isothermal

In within this mentioned material, there is another volume air then I should apply energy equation because I want to see temperature variation in this material and the existence air in material.

[aja1345]

Originally Posted by evcelica

Yes, that is called a Dirichlet boundary condition. But it doesn't sound like a realistic setup to me.

You would apply a volumetric heat source (subdomain) that is a function of local temperature and intended temperature. Cp * Density *(T.intended - T)/1[s] with a source term coefficient of -1[s].

Again, this does not sound realistic if you are intending to simulate heat transfer.

Thanks.



could you please explain more?


what is a source term coefficient of -1[s]? where I should use this in CFX? source coefficient is in w/m^3K in CFX not [s].

[aja1345]

Air in Material have a initial temperature in this problem but air around the material have a Fixed temperature.

[Gert-Jan]

You can do this in Fluent. It has the functionaity to set the temperature in a certain volume to a fixed value. And if something flows through it, it gets that temperature.

It is realistic? Don't think so. But it can be useful. I used it to model the flow through a burner without taking the complete combustion process into account. I just wanted the temperature rise.......

[evcelica]

Originally Posted by aja1345

Thanks.



could you please explain more?


what is a source term coefficient of -1[s]? where I should use this in CFX? source coefficient is in w/m^3K in CFX not [s].

Just make it negative 1[s^-1]*(whatever units it wants to be correct). If your function will change the temperature from local to Intended in 1 second. It just linearizes the source term to make it more numerically stable. It should be around equal to the derivative of the functions effect for the best stability I believe. See the documentation for more details.

If you don't use source term linearization, you can't use a timestep larger than your source terms effect time, or you will get overshoot and numerical instability.