Set temperature t=0
 

Good evening. I am studying the heat transfer between a solid zone and a fluid zone. How can I initialize the temperature of the solid zone?

Thank you for the support!

You can patch temperature value to the domain. See this webpage. The page is for older version but the way it more or less same in newer versions also. You just select variable, zone, input a value and press 'patch'. You can verify by plotting contours for corresponding walls.

Mmmm no.. It doesn't start from that T. I must find another method. Thank you the same.

I didn't catch that. Could you elaborate what exactly you did?

I have a cylinder with an internal diameter with the fluid flow in a ceramic honeycomb (porous medium). The cylinder is a steel vessel, I set coupled as method of heat exchange betweeen the fluid (air) and the solid. I wrote an UDF for the Tinlet, in function of time. If I don't initialize the temperature of the ceramic and the steel it starts from the first value of the UDF. If i fix the temperature of steel and ceramic it is not correct because it means that this temperature will be constant for all the simulation. I don't really know how to solve it, because with patch the temperature deverges and the temperature at t=0 is taken from UDF (higher than the ambient that I'd like to initialize...). It is for my thesis... I don't like Fluent, it is ambiguos. :mad:

I hope you haven't misunderstood patching as fixed temperature. I mean it is not like if you patch 500 K to a domain it will remain 500 K throughout simulation. It will change as your simulation progresses.

It would be interesting to see your case. Can you upload it somewhere?

Nono, I understood the difference. And here is the problem. The size is too big. I don't want to waste your time, I ask you simply... I have two zones, fluid and solid. I don't fix the temperature. What should I do before, do "hybrid initialization" or write values for the patch? In order of time...

I think you do both. Initialize first and then use patch.

Also, if you want the flow to develop first, you can solve only flow equations first. Once the flow is developed then you use patch and then solve flow and energy equations together.

Are you using "compute from" and specifying the inlet? If you use compute from and specify any boundary, Fluent will use the temperature of the boundary as the initial condition. This isn't am ambiguity.

If you know what the initial values are then specify them explicitly using standard initialization.

Patching also works!

If you have divergence errors then it may be because your initial conditions are in an inconsistent state. For example you may think that an initial flow field will be a perfectly uniform temperature but because of the discretization cause by the mesh this is impossible. In this case it is helpful to freeze other variables (like flow) and run a few iterations to get over the numerical error hurdle.

Ops.. I forgot compute from inlet!!! I don't know how to say THANK YOU!!!!! :):):):):)

[QUOTE=LuckyTran;622916]Are you using "compute from" and specifying the inlet? If you use compute from and specify any boundary, Fluent will use the temperature of the boundary as the initial condition. This isn't am ambiguity.

If you know what the initial values are then specify them explicitly using standard initialization.

Compute from is made ''standard initialization'' right?

it is like a transient boundary condition in a sense by making ''compute from'',hovewer, ''specify boundary in hand '' provides it ''constant'' value in all process.

are those right or not?