CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   natural convection in a sealed enclosure (

James March 30, 2001 14:38

natural convection in a sealed enclosure
I am looking at a problem where a number of heating elements are enclosed in a sealed container with helium backfill gas. The problem is one of natural convection from the heater to gas, then gas to container wall. Conduction through the wall and natural convection again to ambient. My question is, can this be addressed as a steady state problem, or is the natural convection inside the container indeterminate?

Selina Tracy March 30, 2001 14:57

Re: natural convection in a sealed enclosure
Before discussing if the flow is steady or not, let's get down to the scaling.

Size of the room(L), Temperatures(T_room, T_outside, T_heater), and what time frame are you interested in.

Where is the heater? Near wall or middle of the room. Once we get these information, we can discuss details.


James March 30, 2001 15:09

Re: natural convection in a sealed enclosure
Ok, we're actually talking nuclear fuel in storage tubes, inside a concrete container. Inside cavity of the container is about 2m x 2m x 4m tall. Lets guess at fuel temperatures of about 150C, outside ambient temperature of 5C. Timescale in years. Hope that helps. Also, what CFD packages would you use for this?


Alton J. Reich March 30, 2001 18:58

Re: natural convection in a sealed enclosure

Nuclear fuel, eh? It takes me back to the days when I was an actual nuclear engineer... But I'm digressing.

I think that if you are talking about time scales of years, then the problem becomes steady state.

I'm going to guess that you're looking at dry storage casks and you need to get a feel for the temperature of the concrete under a variety of conditions. Your inputs would be things like the location of the fuel rods in the cask, thickness of concrete, any structure that connects the fuel to the concrete (like spacers, because conduction is important, too), the heat generation rate from the fuel (probably a design value from a spec), and a maximum ambient temperature. The goal would then be to run a steady state analysis to determine the "equilibrium" temperature of the system (ie: Qgen of the fuel = Qnatural_convection from the outside of the cask)

If you'd like more information on approach, etc. feel free to e-mail me (although I'm going to be travelling next week) or leave me a voice mail message at (256) 726-4919. Also take a look at our web site <>

Regards, Alton

Astrid April 2, 2001 15:48

Re: natural convection in a sealed enclosure
You should consult AEA Technology in the UK too. They provide the CFD-package CFX. AEAT might be interesting for you as they have a strong background in nucleair technology.

Regs, Astrid

All times are GMT -4. The time now is 16:56.