CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   solid with temperatur (http://www.cfd-online.com/Forums/fluent/29761-solid-temperatur.html)

 Konstantin Philipp May 15, 2002 05:39

solid with temperatur

Hi! Got a problem of cooling a solid body(steel!). I have to define the temperature inside the body (grid: default interior, nearly 1100K), and not only the temperature at the walls aoutside the body. With transient simulation I want to calculate the surface reaction. Gimme a hint, please!

 JiChun May 15, 2002 16:19

Re: solid with temperatur

Your problem is a conjugate heat transfer problem. Your computational domain should include both the solid body(steel) and the fluid doamin. In GAMBIT, define these two different zones as solid and fluid zones respectively. I think your calcuation should also include radiation since the temperature is very high. I don't know what kind of help you want. Your question seems unclear to me.

 Konstantin Philipp May 16, 2002 05:01

Re: solid with temperatur

That's cool! The most thing I like on internet is having contact with persons thousands miles away from me! By the way, thank you for your response JiChun! The problem is: how to define uniform temperature (1100K) at the beginning of the simulation inside the body, not only at the surface (wall). The aim is to compute the temperature-distribution inside the solid over time. Maybe this is more information.

 Erwin May 16, 2002 07:19

Re: solid with temperatur

Define a small portion of your zone in Gambit as a separate solid zone, call it for example 'core'. Then in Fluent you can set the initial value of core to be at whatever temperature. You can use either the 'initialize' or 'patch' command. Especially patching the zones might be useful for you.

 JiChun May 16, 2002 15:29

Re: solid with temperatur

That's too easy, just do as what I said, then in FLUENT select the solid zone and patch a uniform temperature 1100K to this region. That's it.

 David Shkval May 22, 2002 09:52

One more question

Excuse me, here is one more question:

As appear in transient conjugate problems, we shall simulate unsteady history of the heat transfering. Here comes "rigid" problem when time scale in fluid zones compared to that in solid zones. e.g., as limited by the stability of solver in fluid zones, each real time step may uplimited to 10^4, while time step in solid zones can be much more larger than this value.

My problem is: when simulate hypersonic conjugate heat transfer problem, since up-limits of time step in solid&fluid zones are different(maybe 10^4 : 1), which solver shall we use in Fluent?

1. explicit scheme: convergence is difficult and costly. 2. implicit scheme with pre-condictioning method: dunno if it's suitable in hypersonic case.

 Alammar May 23, 2002 13:20

Re: One more question

Your best bet is to use the coupled, implicit solver. Theoretically, it's unconditionally stable; i.e., time step is not a problem. However, you might develop some instabilities due to, for example, high gradients of shocks and reactions; but then you can relax your scheme temporarly with the Courant number. If you use the segregate solver, you're gonna end up under-relaxing your equations big time!

 David Shkval May 24, 2002 04:46

Re: One more question

thank u Alammar, I'll try

 All times are GMT -4. The time now is 21:25.