Natural Convection heat transfer problem

srinivasa · February 1, 2012, 07:55

Hello everyone.
I have one problem related to heat transfer natural convection.
Problem statement: A rectangular box containing bus bars,due to flow of current heat is generated.The heat is distributed by mode of convection(natural Convection).Intially there are openings below and above of the rectangular box which acts as a inlet and outlet for airflow.I simualted it in fluent and i got the satisfactory result.Now,i have to simulate the similar one with NO INLET and OUTLET that is the rectangular box to be fully enlcosed.I have to find the bus bar temperatures now with the air which is intially have.

can any one help me to give the boundary conditions .

regards
srinivasa

ANTONY A R V · February 1, 2012, 08:39

You can take the box domain as fluid and the outer wall as insulated(heat flux zero). The heat released by the heater surface can be simulared by giving the heat flux boundary condition. Heat provided by the heater can be converted to heat flux by dividing the heat by the surface area of the heater.

Enable the energy.

Natural convection can be simulated by enabling the gravity in the operating condition panel and also giving the gravity value correctly in the gravitaational accelaration form.(careful about the direction of g).

srinivasa · February 1, 2012, 08:52

thank u..
i will try for it

srinivasa · February 1, 2012, 09:11

Hi Antony..

were i have to define the air temperature.

ANTONY A R V · February 1, 2012, 09:42

Give the initial condition if it is transient analysis and intial solution if it is steady state analysis in the iterate panel (may be 30 deg cent) .

kat479 · February 2, 2012, 00:47

hey whenever i give heat flux to the source and source shadow.its showing divergence detected in AMG Solver.pliz help.One more thing since its a electric current heat dissipation should we have to give heat generation??

srinivasa · February 2, 2012, 02:16

Hi antony.
i have given air temp as 24 at initalisation. as my problem is steady state.. but it solution is not converging.shall i use any relaxation factors.

kat479 · February 2, 2012, 02:57

are you working for a switchgear? internal arc modelling??

ANTONY A R V · February 2, 2012, 19:35

Dear mr Srinivasa

Tell me whether your monitor graph residuals are decreasing or increasing with iterations.

If the slope of the decrease is very less, you can go for relaxation.

Also you can use boussniques model for fluid density (instead of constant) to model the change in density with temperature of the fluid in the fluid material property panel.

ANTONY A R V · February 2, 2012, 19:46

Dear Mr Karthik

If you are modelling the heater as surface(face) You can give surface heat flux bc.

If you are modelling it as a solid volume, you can give it as heat generation per unit volume.

If it is a volume there should be mesh connectivity between heater and surrounding fluid.

If divergence is there in temperature it may be due to wrong bc, initial condition or check the 'solve from' in the iterate panel. You can change the solve from location.

You can also use the boussnisques model for density(details in previous message)

gaurav_1guru · February 3, 2012, 15:09

Dear Friends,
I am too working on the same problem.
Please give me any details u have regarding the problem because i'm too stuck with the solution.
is it necessary to initialize the solution from a specific zone.

please reply as soon as possible.. my email id is gaurav_1guru@rediff.com

kat479 · February 5, 2012, 22:31

exactly same problem im facing now

srinivasa · February 6, 2012, 03:05

Dear Antony..

My Residual graphs are osciallting.. i am using the Boussinesq model only.

ANTONY A R V · February 6, 2012, 09:27

Dear Mr Srinivasa

Are you doing transient analysis. Can you upload the picture of residual graph. I will try giving suggestions.

ANTONY A R V · February 6, 2012, 09:29

Dear Mr Gaurav

Please eloborate about your problem for giving suggestions.

gaurav_1guru · February 6, 2012, 12:28

http://www.cfd-online.com/Forums/flu...ng-fluent.html

please go through this link it has description of geometry and the problem..

kat479 · February 6, 2012, 22:24

from which zone should we do solution initialisation??

gaurav_1guru · February 7, 2012, 11:00

I have attached the residuals plot for steady analysis...
can u suggest me how to do transient analysis..

kat479 · February 8, 2012, 03:50

which solver should i select for transient condition natural convection involving high temperature and pressure variation...shock waves invovled adaibatic change in an enclosure

gaurav_1guru · February 17, 2012, 10:43

please reply soon

February 1, 2012, 07:55	Natural Convection heat transfer problem	#1
srinivasa New Member sree Join Date: Jun 2011 Posts: 12 Rep Power: 14	Hello everyone. I have one problem related to heat transfer natural convection. Problem statement: A rectangular box containing bus bars,due to flow of current heat is generated.The heat is distributed by mode of convection(natural Convection).Intially there are openings below and above of the rectangular box which acts as a inlet and outlet for airflow.I simualted it in fluent and i got the satisfactory result.Now,i have to simulate the similar one with NO INLET and OUTLET that is the rectangular box to be fully enlcosed.I have to find the bus bar temperatures now with the air which is intially have. can any one help me to give the boundary conditions . regards srinivasa

February 1, 2012, 08:39		#2
ANTONY A R V New Member ANTONY AROUL RAJ V Join Date: Jan 2012 Location: Chennai Posts: 22 Rep Power: 14	You can take the box domain as fluid and the outer wall as insulated(heat flux zero). The heat released by the heater surface can be simulared by giving the heat flux boundary condition. Heat provided by the heater can be converted to heat flux by dividing the heat by the surface area of the heater. Enable the energy. Natural convection can be simulated by enabling the gravity in the operating condition panel and also giving the gravity value correctly in the gravitaational accelaration form.(careful about the direction of g). kat479 likes this.

February 2, 2012, 19:46		#10
ANTONY A R V New Member ANTONY AROUL RAJ V Join Date: Jan 2012 Location: Chennai Posts: 22 Rep Power: 14	Dear Mr Karthik If you are modelling the heater as surface(face) You can give surface heat flux bc. If you are modelling it as a solid volume, you can give it as heat generation per unit volume. If it is a volume there should be mesh connectivity between heater and surrounding fluid. If divergence is there in temperature it may be due to wrong bc, initial condition or check the 'solve from' in the iterate panel. You can change the solve from location. You can also use the boussnisques model for density(details in previous message) kat479 likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Natural convection in a closed domain STILL NEEDING help!	Yr0gErG	FLUENT	4	December 2, 2019 00:04
Multiregion Heat Transfer + natural convection (water) with chtMultiRegionFoam	fattychickenrun	OpenFOAM	5	October 31, 2011 16:53
natural convection	mehrdadeng	CFX	10	February 25, 2011 05:25
Natural Convection with heat generation	krishnachandranr	Main CFD Forum	0	July 28, 2009 04:22
Heat transfer problem	Brian	CFX	0	September 13, 2004 01:19

February 1, 2012, 08:52		#3
srinivasa New Member sree Join Date: Jun 2011 Posts: 12 Rep Power: 14	thank u.. i will try for it

February 1, 2012, 09:11		#4
srinivasa New Member sree Join Date: Jun 2011 Posts: 12 Rep Power: 14	Hi Antony.. were i have to define the air temperature.

February 1, 2012, 09:42		#5
ANTONY A R V New Member ANTONY AROUL RAJ V Join Date: Jan 2012 Location: Chennai Posts: 22 Rep Power: 14	Give the initial condition if it is transient analysis and intial solution if it is steady state analysis in the iterate panel (may be 30 deg cent) .

February 2, 2012, 00:47		#6
kat479 New Member kat Join Date: Jan 2012 Posts: 8 Rep Power: 14	hey whenever i give heat flux to the source and source shadow.its showing divergence detected in AMG Solver.pliz help.One more thing since its a electric current heat dissipation should we have to give heat generation??

February 2, 2012, 02:16		#7
srinivasa New Member sree Join Date: Jun 2011 Posts: 12 Rep Power: 14	Hi antony. i have given air temp as 24 at initalisation. as my problem is steady state.. but it solution is not converging.shall i use any relaxation factors.

February 2, 2012, 02:57		#8
kat479 New Member kat Join Date: Jan 2012 Posts: 8 Rep Power: 14	are you working for a switchgear? internal arc modelling??

February 2, 2012, 19:35		#9
ANTONY A R V New Member ANTONY AROUL RAJ V Join Date: Jan 2012 Location: Chennai Posts: 22 Rep Power: 14	Dear mr Srinivasa Tell me whether your monitor graph residuals are decreasing or increasing with iterations. If the slope of the decrease is very less, you can go for relaxation. Also you can use boussniques model for fluid density (instead of constant) to model the change in density with temperature of the fluid in the fluid material property panel.

February 3, 2012, 15:09		#11
gaurav_1guru New Member guru Join Date: Jan 2012 Posts: 9 Rep Power: 14	Dear Friends, I am too working on the same problem. Please give me any details u have regarding the problem because i'm too stuck with the solution. is it necessary to initialize the solution from a specific zone. please reply as soon as possible.. my email id is gaurav_1guru@rediff.com

February 5, 2012, 22:31		#12
kat479 New Member kat Join Date: Jan 2012 Posts: 8 Rep Power: 14	exactly same problem im facing now

February 6, 2012, 03:05		#13
srinivasa New Member sree Join Date: Jun 2011 Posts: 12 Rep Power: 14	Dear Antony.. My Residual graphs are osciallting.. i am using the Boussinesq model only.

February 6, 2012, 09:27		#14
ANTONY A R V New Member ANTONY AROUL RAJ V Join Date: Jan 2012 Location: Chennai Posts: 22 Rep Power: 14	Dear Mr Srinivasa Are you doing transient analysis. Can you upload the picture of residual graph. I will try giving suggestions.

February 6, 2012, 09:29		#15
ANTONY A R V New Member ANTONY AROUL RAJ V Join Date: Jan 2012 Location: Chennai Posts: 22 Rep Power: 14	Dear Mr Gaurav Please eloborate about your problem for giving suggestions.

February 6, 2012, 12:28		#16
gaurav_1guru New Member guru Join Date: Jan 2012 Posts: 9 Rep Power: 14	http://www.cfd-online.com/Forums/flu...ng-fluent.html please go through this link it has description of geometry and the problem..

February 6, 2012, 22:24		#17
kat479 New Member kat Join Date: Jan 2012 Posts: 8 Rep Power: 14	from which zone should we do solution initialisation??

February 8, 2012, 03:50		#19
kat479 New Member kat Join Date: Jan 2012 Posts: 8 Rep Power: 14	which solver should i select for transient condition natural convection involving high temperature and pressure variation...shock waves invovled adaibatic change in an enclosure

February 17, 2012, 10:43		#20
gaurav_1guru New Member guru Join Date: Jan 2012 Posts: 9 Rep Power: 14	please reply soon