Couple outlet to inlet to achieve a fully developed flow.

raghu.tejaswi · May 2, 2012, 06:57

Hello FLUENT users,

I am trying to simulate a simple 2D FULLY DEVELOPED flow (for both temp and velocity) between parallel plates in steady state. The working medium is water. To achieve a fully developed flow, the length of the domain becomes too long according to equations for turbulent entry lengths. Hence I would like to couple the values from outlet to inlet and solve both the flow and energy equations until convergence is reached for a short length of the domain.

For this I would like to write a UDF. I am very new to UDfs. My idea is to initialise velocity and temp in a udf. Use the value of velocity and temp from outlet and set it as inlet. This needs to continue until convergence is reached.

I understand periodic BC does exactly the above for flow equations. I am not sure if I can have a thermally developed flow too with periodic BC.

I just need to know how I can map the outlet surface to inlet in FLUENT using udf.

Blader · May 3, 2012, 15:21

Hi,

I have the same problem. I need to link outlet and inlet. Can anyone help us ?

Thanks in advance

raghu.tejaswi · May 4, 2012, 03:06

Hi Blader,

I think I have found a solution. Use periodic BC in FLUENT. But pls make sure you meet all the necessary criteria before beginning the simulation. Run the simulation till residuals converge (normally one would monitor the outlet temp but in periodic BC case, the outlet is a shadow of the inlet..)

Pls try this for your case and do let me know if you are successful.

ciao..
Raghu

Blader · May 4, 2012, 03:55

I will try to do the simulation with periodic BC and I will comment the results.

Thank you very much.

raghu.tejaswi · May 10, 2012, 04:30

Hi Blader,

I hope periodic BC worked for you coz it worked for me pretty well. The flow profile is fully developed and the conditions to attain the fully developed thermal profile is achieved..

My problem is solved, atleast temporarily...

Raghu

Blader · May 10, 2012, 05:38

Hello,

In my case doesn't work very well.

I make a brief summary of my problem:

I want to simulate a closed loop pipe. I don't found a way to do it. I decided to do it with a open pipe linking the output to input. However, I just want to link the temperature of the output to input.

It really works with periodic BC properties?

thanks in advance !

raghu.tejaswi · May 10, 2012, 06:31

The concept of Periodic BC used by me is to achieve a fully developed flow. Since I use water as the working medium (whose Pr = 7.01), theoretically the entrance length is very high. I need a huge computational domain to achieve this. Instead I couple inlet to outlet to achieve. I need both velocity and temp to achieve fully developed condition. But here the outlet parameters are made as the inlet and this you see creates a loop...

http://hpce.iitm.ac.in/website/Manua...f/ug/chp13.pdf

Im really not sure about periodic BC in your case.

claudia.h · July 18, 2013, 09:49

Hello,

i am looking for a way to couple the outlet with the inlet because i want to use the outlet parameters (velocity, temperature) for the inlet boundary for each iteration. Then i found your posting and hoped to get an input how to solve such a problem but unfortunately the link doesn´t work. It would be a great help for me if you could give me an hint how to solve such a problem.

Thanks in advance !

raghu.tejaswi · July 18, 2013, 10:04

Hi Claudia,

Did you try periodic BC?? It should work. This is how I got my results and they seemed to match quite good with the theoretical values and I was also able to give a Logical Explanation.

http://cdlab2.fluid.tuwien.ac.at/LEH...ug/node260.htm

I was looking at the second condition explained here. Are you trying to achieve the same? The reason for choosing the periodic BC is explained in my previous Posts.

cheers
raghu

May 2, 2012, 06:57	Couple outlet to inlet to achieve a fully developed flow.	#1
raghu.tejaswi New Member Learner Join Date: Nov 2011 Location: Ingolstadt Posts: 27 Rep Power: 15	Hello FLUENT users, I am trying to simulate a simple 2D FULLY DEVELOPED flow (for both temp and velocity) between parallel plates in steady state. The working medium is water. To achieve a fully developed flow, the length of the domain becomes too long according to equations for turbulent entry lengths. Hence I would like to couple the values from outlet to inlet and solve both the flow and energy equations until convergence is reached for a short length of the domain. For this I would like to write a UDF. I am very new to UDfs. My idea is to initialise velocity and temp in a udf. Use the value of velocity and temp from outlet and set it as inlet. This needs to continue until convergence is reached. I understand periodic BC does exactly the above for flow equations. I am not sure if I can have a thermally developed flow too with periodic BC. I just need to know how I can map the outlet surface to inlet in FLUENT using udf.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Nusselt numbers for fully developed laminar flow	Danro	Siemens	8	July 19, 2012 08:22
Fully developed flow at inlet ?	sezenempire	FLUENT	0	March 27, 2012 06:37
total pressure boundary problem ==> flow from outlet to inlet!!	mrshb4	OpenFOAM	0	November 20, 2010 13:41
Writing an expression for fully developed flow!	Usman	CFX	12	December 20, 2007 12:26
what the result is negatif pressure at inlet	chong chee nan	FLUENT	0	December 29, 2001 06:13

May 3, 2012, 15:21		#2
Blader New Member blai Join Date: Mar 2012 Posts: 8 Rep Power: 14	Hi, I have the same problem. I need to link outlet and inlet. Can anyone help us ? Thanks in advance

May 4, 2012, 03:06		#3
raghu.tejaswi New Member Learner Join Date: Nov 2011 Location: Ingolstadt Posts: 27 Rep Power: 15	Hi Blader, I think I have found a solution. Use periodic BC in FLUENT. But pls make sure you meet all the necessary criteria before beginning the simulation. Run the simulation till residuals converge (normally one would monitor the outlet temp but in periodic BC case, the outlet is a shadow of the inlet..) Pls try this for your case and do let me know if you are successful. ciao.. Raghu

May 4, 2012, 03:55		#4
Blader New Member blai Join Date: Mar 2012 Posts: 8 Rep Power: 14	I will try to do the simulation with periodic BC and I will comment the results. Thank you very much.

May 10, 2012, 04:30		#5
raghu.tejaswi New Member Learner Join Date: Nov 2011 Location: Ingolstadt Posts: 27 Rep Power: 15	Hi Blader, I hope periodic BC worked for you coz it worked for me pretty well. The flow profile is fully developed and the conditions to attain the fully developed thermal profile is achieved.. My problem is solved, atleast temporarily... Raghu

May 10, 2012, 05:38		#6
Blader New Member blai Join Date: Mar 2012 Posts: 8 Rep Power: 14	Hello, In my case doesn't work very well. I make a brief summary of my problem: I want to simulate a closed loop pipe. I don't found a way to do it. I decided to do it with a open pipe linking the output to input. However, I just want to link the temperature of the output to input. It really works with periodic BC properties? thanks in advance !

May 10, 2012, 06:31		#7
raghu.tejaswi New Member Learner Join Date: Nov 2011 Location: Ingolstadt Posts: 27 Rep Power: 15	The concept of Periodic BC used by me is to achieve a fully developed flow. Since I use water as the working medium (whose Pr = 7.01), theoretically the entrance length is very high. I need a huge computational domain to achieve this. Instead I couple inlet to outlet to achieve. I need both velocity and temp to achieve fully developed condition. But here the outlet parameters are made as the inlet and this you see creates a loop... http://hpce.iitm.ac.in/website/Manua...f/ug/chp13.pdf Im really not sure about periodic BC in your case.

July 18, 2013, 09:49		#8
claudia.h New Member Claudia Hintringer Join Date: May 2010 Posts: 10 Rep Power: 16	Hello, i am looking for a way to couple the outlet with the inlet because i want to use the outlet parameters (velocity, temperature) for the inlet boundary for each iteration. Then i found your posting and hoped to get an input how to solve such a problem but unfortunately the link doesn´t work. It would be a great help for me if you could give me an hint how to solve such a problem. Thanks in advance !

July 18, 2013, 10:04		#9
raghu.tejaswi New Member Learner Join Date: Nov 2011 Location: Ingolstadt Posts: 27 Rep Power: 15	Hi Claudia, Did you try periodic BC?? It should work. This is how I got my results and they seemed to match quite good with the theoretical values and I was also able to give a Logical Explanation. http://cdlab2.fluid.tuwien.ac.at/LEH...ug/node260.htm I was looking at the second condition explained here. Are you trying to achieve the same? The reason for choosing the periodic BC is explained in my previous Posts. cheers raghu