Pressure Inlet BC - Transient run

narendra.gadwal · March 7, 2012, 03:02

Hello,

I have a tank in which initially air is present at atmospheric pressure.
I have an inlet from which I am pumping air at 6bar, 8bar and 12 bar pressure for 3 cases.
My aim is to find out the time taken for the pressure raise inside the tank as of the inlet pressure.
I gave pressure inlet BC with the values normal to the boundary and I am monitoring the area average of the static pressure of the tank surface w.r.t flow time.
The curve settles down too fast as compared to the experimental values. Is there anything else I need to do ?

Thanks in advance !

duri · March 9, 2012, 15:38

Is your initial flow rate or static pressure is same as experiment?

narendra.gadwal · March 10, 2012, 06:05

yes ... the gauge pressure at the inlet is 14bar ..

faw · March 11, 2012, 15:36

Hi, I would like to know have you tried using different time step size? Or what solver are you using? (SIMPLE/PISO) Different viscous model also will affect the settling time as well.

My project is similar as yours, but using oil as the fluid. I have done some comparisons using different solver, viscous models and changing time step size as well, and it seems that the results are different. But I haven't compared it with the experiment value yet, maybe by next week. Then maybe we can share our thoughts

-faw-

narendra.gadwal · March 11, 2012, 23:27

Hello,

My mesh size is 500k tetra and my time step = 0.001sec.
I am using SIMPLE scheme ..

Thanks !

faw · March 12, 2012, 10:21

Have you tried using higher time step size? Apparantly in my case, when I'm using larger time step size, the curve settles down slower (longer time taken)

narendra.gadwal · March 12, 2012, 11:59

Don't you think the result should be independent of time step size ?

faw · March 12, 2012, 12:29

Actually it is supposed to be like that, but when I tried using different time step size, the settling time did change a lot and that made me confused actually

duri · March 13, 2012, 09:18

Hi,

Time taken should be independent of time step. I strongly suspect on the initial condition of the problem. Is inital flow rate is correct. In experiments flow starts from zero and reaches its peak value and then drops as pressure difference goes down. It is very difficult to simulate this exactly. If flow rise time is negligible then it can be ignored but peak flow rate has to be used to initialize. (the term peak used here could be misleading)

March 7, 2012, 03:02	Pressure Inlet BC - Transient run	#1
narendra.gadwal Member Narendra Gadwal Join Date: Jul 2010 Location: India Posts: 35 Rep Power: 16	Hello, I have a tank in which initially air is present at atmospheric pressure. I have an inlet from which I am pumping air at 6bar, 8bar and 12 bar pressure for 3 cases. My aim is to find out the time taken for the pressure raise inside the tank as of the inlet pressure. I gave pressure inlet BC with the values normal to the boundary and I am monitoring the area average of the static pressure of the tank surface w.r.t flow time. The curve settles down too fast as compared to the experimental values. Is there anything else I need to do ? Thanks in advance !

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Pressure Outlet Guage pressure	Mohsin	FLUENT	36	April 29, 2016 18:16
Mass flow inlet and pressure outlet issue	nikhil	FLUENT	5	December 11, 2013 13:30
Pressure Inlet BC	Prasad	FLUENT	3	January 4, 2009 23:55
How to achieve constant static pressure in inlet?	Zou Jianfeng	FLUENT	0	October 21, 2008 02:55
Setting dynamic pressure as the inlet BC	ruined	FLUENT	1	March 26, 2008 06:30

March 9, 2012, 15:38		#2
duri Senior Member duri Join Date: May 2010 Posts: 245 Rep Power: 17	Is your initial flow rate or static pressure is same as experiment?

March 10, 2012, 06:05		#3
narendra.gadwal Member Narendra Gadwal Join Date: Jul 2010 Location: India Posts: 35 Rep Power: 16	yes ... the gauge pressure at the inlet is 14bar ..

March 11, 2012, 15:36		#4
faw New Member Fara Join Date: Feb 2012 Posts: 6 Rep Power: 14	Hi, I would like to know have you tried using different time step size? Or what solver are you using? (SIMPLE/PISO) Different viscous model also will affect the settling time as well. My project is similar as yours, but using oil as the fluid. I have done some comparisons using different solver, viscous models and changing time step size as well, and it seems that the results are different. But I haven't compared it with the experiment value yet, maybe by next week. Then maybe we can share our thoughts -faw-

March 11, 2012, 23:27		#5
narendra.gadwal Member Narendra Gadwal Join Date: Jul 2010 Location: India Posts: 35 Rep Power: 16	Hello, My mesh size is 500k tetra and my time step = 0.001sec. I am using SIMPLE scheme .. Thanks !

March 12, 2012, 10:21		#6
faw New Member Fara Join Date: Feb 2012 Posts: 6 Rep Power: 14	Have you tried using higher time step size? Apparantly in my case, when I'm using larger time step size, the curve settles down slower (longer time taken)

March 12, 2012, 11:59		#7
narendra.gadwal Member Narendra Gadwal Join Date: Jul 2010 Location: India Posts: 35 Rep Power: 16	Don't you think the result should be independent of time step size ?

March 12, 2012, 12:29		#8
faw New Member Fara Join Date: Feb 2012 Posts: 6 Rep Power: 14	Actually it is supposed to be like that, but when I tried using different time step size, the settling time did change a lot and that made me confused actually

March 13, 2012, 09:18		#9
duri Senior Member duri Join Date: May 2010 Posts: 245 Rep Power: 17	Hi, Time taken should be independent of time step. I strongly suspect on the initial condition of the problem. Is inital flow rate is correct. In experiments flow starts from zero and reaches its peak value and then drops as pressure difference goes down. It is very difficult to simulate this exactly. If flow rise time is negligible then it can be ignored but peak flow rate has to be used to initialize. (the term peak used here could be misleading)