natural convection in stratified zone

cfdlame · November 12, 2011, 14:11

I'm using Fluent forum, tell me if I've choosen incorrectly.

The problem is: For different solver setting I'm experiencing natural convection in zone, where initialized temperature gradient is less than the lapse rate - convection should't occure.

I'm investigating convection in closed cube 3kmx3kmx2km with adiabatic boundary condition (hex grid 150000 elements). Initialized temperature gradient is decreasing 0,5K/100m (adiabatic lapse rate is 1K/100m). I've tried: boussinesq, ideal gas, and real gas for a density model, for both solvers: density, and pressure based (presto, second order for everything). I don't even know solver type matters, probably not. UDF is written fine. In every case convection occured. Somebody have any idea?

Tomek

cfdlame · November 14, 2011, 05:08

any ideas? maybe I should change forum?

sandeep_tu · November 14, 2011, 11:53

I can make some suggestions,

* May be don´t use Gravitional acceleration.

* Density is very important in FLUENT. You should choose this based up on your problem. May be this helps,

1. For compressible flows, the ideal gas law is the appropriate density relationship.

2. For incompressible flows, you may choose constant density .

all other options are used to model natural convection (for example Boussinesq model)

cfdlame · November 14, 2011, 16:00

This is not exactly answer to my question, but thanks.

Once again: In my model convection occurs, but it should't. All properieties for convection should be enabled (ideal gas, gravity, energy equation etc.) but it should not happen, because tempreture gradient is too small (below adiabatic lapse rate)

November 12, 2011, 14:11	natural convection in stratified zone	#1
cfdlame New Member Join Date: Nov 2011 Location: Poland Posts: 7 Rep Power: 15	I'm using Fluent forum, tell me if I've choosen incorrectly. The problem is: For different solver setting I'm experiencing natural convection in zone, where initialized temperature gradient is less than the lapse rate - convection should't occure. I'm investigating convection in closed cube 3kmx3kmx2km with adiabatic boundary condition (hex grid 150000 elements). Initialized temperature gradient is decreasing 0,5K/100m (adiabatic lapse rate is 1K/100m). I've tried: boussinesq, ideal gas, and real gas for a density model, for both solvers: density, and pressure based (presto, second order for everything). I don't even know solver type matters, probably not. UDF is written fine. In every case convection occured. Somebody have any idea? Tomek

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
natural convection problem with radiation	jorien	CFX	0	October 14, 2011 10:26
Problem in running ICEM grid in Openfoam	Tarak	OpenFOAM	6	September 9, 2011 18:51
Problem in IMPORT of ICEM input file in FLUENT	csvirume	FLUENT	2	September 9, 2009 02:08
BC for Natural Convection in Stratified environment	toyoto	Main CFD Forum	0	September 3, 2009 00:10
Modelling natural convection in stratified environment	toyoto	FLUENT	0	September 3, 2009 00:05

November 14, 2011, 05:08		#2
cfdlame New Member Join Date: Nov 2011 Location: Poland Posts: 7 Rep Power: 15	any ideas? maybe I should change forum?

November 14, 2011, 11:53		#3
sandeep_tu Member Sandeep Join Date: Apr 2009 Location: Munich, Germany Posts: 30 Rep Power: 17	I can make some suggestions, * May be don´t use Gravitional acceleration. * Density is very important in FLUENT. You should choose this based up on your problem. May be this helps, 1. For compressible flows, the ideal gas law is the appropriate density relationship. 2. For incompressible flows, you may choose constant density . all other options are used to model natural convection (for example Boussinesq model)

November 14, 2011, 16:00		#4
cfdlame New Member Join Date: Nov 2011 Location: Poland Posts: 7 Rep Power: 15	This is not exactly answer to my question, but thanks. Once again: In my model convection occurs, but it should't. All properieties for convection should be enabled (ideal gas, gravity, energy equation etc.) but it should not happen, because tempreture gradient is too small (below adiabatic lapse rate)