Problem in laminar natural convection in cavity

realfear · June 26, 2014, 07:06

hi every one
i have modeled a 2d cavity in fluent and openfoam for natural convection heat transfer
they gave me the same results but not the same with the experimental results.
so i'm asking for your advices
the upper edge in cavity is adiabatic and left and right wall are in constant temp (Tc) and a part of lower wall is in constant temp(Th) and some part of it are adiabatic
i have attached the steps i used in modeling the problem in fluent and the results as well
i really need your helps because i;m about to fail my coarse!!!

thank you in advance
Nuextent.xlsx

2.JPG

3.JPG

4.JPG

5.JPG

realfear · June 26, 2014, 07:08

the other steps

6.JPG

7.JPG

8.JPG

123.JPG

1234.JPG

Mohawk · June 26, 2014, 09:44

From density choose boussniq

realfear · June 26, 2014, 10:09

Quote:

Originally Posted by Mohawk

From density choose boussniq

i did but it converged with 3 iterations!!!!!!!!!!!!!!!!!!!!!

realfear · June 27, 2014, 06:10

anyone ????
please
i need help!!

kad · June 27, 2014, 10:55

Make sure that Nusselt number is calculated the same way in simulation and experiment. Nusselt number in fluent is calculated as,

$Nu = \frac{h_{eff} \cdot L_{ref}}{k}$
with
$h_{eff} = \frac{q}{T_{wall}-T_{ref}}$ .

q is the local heat flux, k is the local heat conductivity. The reference values are defined in the corresponding panel "reference values" in fluent (Length and Temperature).

realfear · June 27, 2014, 11:47

Quote:

Originally Posted by kad

Make sure that Nusselt number is calculated the same way in simulation and experiment. Nusselt number in fluent is calculated as,

$Nu = \frac{h_{eff} \cdot L_{ref}}{k}$
with
$h_{eff} = \frac{q}{T_{wall}-T_{ref}}$ .

q is the local heat flux, k is the local heat conductivity. The reference values are defined in the corresponding panel "reference values" in fluent (Length and Temperature).

tanx for responding
i used Forward finite difference scheme for calculating Nu
Nu was defined as d(teta)/dY
teta is the dimensionless temperature
Y is the dimensionless lenght

June 26, 2014, 07:06	Problem in laminar natural convection in cavity	#1
realfear New Member Join Date: Mar 2014 Posts: 16 Rep Power: 12	hi every one i have modeled a 2d cavity in fluent and openfoam for natural convection heat transfer they gave me the same results but not the same with the experimental results. so i'm asking for your advices the upper edge in cavity is adiabatic and left and right wall are in constant temp (Tc) and a part of lower wall is in constant temp(Th) and some part of it are adiabatic i have attached the steps i used in modeling the problem in fluent and the results as well i really need your helps because i;m about to fail my coarse!!! thank you in advance Nuextent.xlsx 2.JPG 3.JPG 4.JPG 5.JPG

June 27, 2014, 06:10	asking for help	#5
realfear New Member Join Date: Mar 2014 Posts: 16 Rep Power: 12	anyone ???? please i need help!!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Natural convection convergence problem	Bruno Machado	FLUENT	10	January 31, 2018 05:18
Problem with natural convection	claudia.h	FLUENT	2	July 3, 2012 04:10
transient simulation: natural convection problem?	Basics	CFX	3	September 25, 2002 10:42
ON TURBULET MODEL FOR A NATURAL CONVECTION PROBLEM	varghese	FLUENT	6	February 6, 2002 06:23
Natural Convection Problem	Sridhar	FLUENT	2	September 13, 2001 16:16

June 26, 2014, 07:08		#2
realfear New Member Join Date: Mar 2014 Posts: 16 Rep Power: 12	the other steps 6.JPG 7.JPG 8.JPG 123.JPG 1234.JPG

June 26, 2014, 09:44		#3
Mohawk Member Mustafa Join Date: May 2013 Posts: 54 Rep Power: 13	From density choose boussniq

June 27, 2014, 10:55		#6
kad Senior Member Join Date: Feb 2013 Location: Germany Posts: 200 Rep Power: 24	Make sure that Nusselt number is calculated the same way in simulation and experiment. Nusselt number in fluent is calculated as, $Nu = \frac{h_{eff} \cdot L_{ref}}{k}$ with $h_{eff} = \frac{q}{T_{wall}-T_{ref}}$ . q is the local heat flux, k is the local heat conductivity. The reference values are defined in the corresponding panel "reference values" in fluent (Length and Temperature).