CFD-Post, expression for pressure fluctuations

hand90 · August 29, 2018, 06:59

Hi everyone

I have a transient simulation of a simple pipe flow with a pressure gradient. I am interested in the pressure fluctuation. The pressure fluctuation is defined by:
p'=p-p_mean

To evaluate the mean pressure I selected the arithmetic pressure average in CFX-Pre which was successful evaluated.
I then created an expression "Pressure-Pressure.Trnavg". However I realized that "Pressure.Trnavg" is only accessible at the Full-Final time step. This prevents me to look at the pressure fluctuation at any specific time step.

Does anyone know how I can change the expression so that Pressure.Trnavg is always at the final time step giving me the correct pressure fluctuation?

Your help is much appreciated!

ghorrocks · August 29, 2018, 08:10

I would recommend doing post-processing like this outside of CFX. Use a monitor point to export the data, then do your post processing in matlab, python, excel or whatever you like.

hand90 · August 29, 2018, 08:17

Thank you for your quick responds!

I have been using monitor points and then post process the data in excel and matlab. However monitor points only gives me a snapshot of what is happening at a certain point in the domain while I am trying to see how the pressure fluctuation propagates through the entire domain (2D domain). Currently I am considering looking into ParaView but was hoping that there is a solution for this in CFD Post.

Any idea?

ghorrocks · August 29, 2018, 08:34

Here's a work around. I am not certain it works, but worth a try: Try defining a new variable, defined as Pressure.Trnavg. Make sure you are at the final time step so the variable is defined. Then go back to a previous time step and hopefully the new variable keeps its values, and then you can evaluate p-(New Variable).

Gert-Jan · August 29, 2018, 09:34

Alternative:
- export the Pressure.Trnavg field as 3D-csv.
- import it into Pre and assign it to an additional variable.
- continue the calculation for one period, while saving pressure and the add.variable to the required transient results files.

This will give you the actual pressure and the average pressure at each timestep, allowing you to calculate the difference.

hand90 · August 29, 2018, 11:39

I can confirm that ghorrocks method worked.

1. Choose the correct time step
2. Create a new variable (Method: Frozen Copy)
3. The new variable can then be used in any expression at any time step, keeping the "Frozen" value.

So far I have not tried the second method suggested but I may try it in the future.

Thank you for your help

August 29, 2018, 06:59	CFD-Post, expression for pressure fluctuations	#1
hand90 Member Thomas Join Date: Nov 2017 Posts: 37 Rep Power: 9	Hi everyone I have a transient simulation of a simple pipe flow with a pressure gradient. I am interested in the pressure fluctuation. The pressure fluctuation is defined by: p'=p-p_mean To evaluate the mean pressure I selected the arithmetic pressure average in CFX-Pre which was successful evaluated. I then created an expression "Pressure-Pressure.Trnavg". However I realized that "Pressure.Trnavg" is only accessible at the Full-Final time step. This prevents me to look at the pressure fluctuation at any specific time step. Does anyone know how I can change the expression so that Pressure.Trnavg is always at the final time step giving me the correct pressure fluctuation? Your help is much appreciated!

August 29, 2018, 08:10		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	I would recommend doing post-processing like this outside of CFX. Use a monitor point to export the data, then do your post processing in matlab, python, excel or whatever you like. __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

August 29, 2018, 08:34		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	Here's a work around. I am not certain it works, but worth a try: Try defining a new variable, defined as Pressure.Trnavg. Make sure you are at the final time step so the variable is defined. Then go back to a previous time step and hopefully the new variable keeps its values, and then you can evaluate p-(New Variable). __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

August 29, 2018, 09:34		#5
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,928 Rep Power: 28	Alternative: - export the Pressure.Trnavg field as 3D-csv. - import it into Pre and assign it to an additional variable. - continue the calculation for one period, while saving pressure and the add.variable to the required transient results files. This will give you the actual pressure and the average pressure at each timestep, allowing you to calculate the difference. Last edited by Gert-Jan; August 29, 2018 at 19:03.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Using CFD Post for OpenFoam results	Karpfen	OpenFOAM Post-Processing	3	January 19, 2018 09:48
couldn't find the cst.file in CFD post folder	Haonan Uwin	FLUENT	0	January 10, 2018 02:25
On the CFD market and trends	sbaffini	Main CFD Forum	14	June 13, 2017 12:48
Post-processing star ccm+ results in Ansys CFD Post	sidharath	STAR-CCM+	4	April 10, 2017 12:49
Automated post processing using CFD Post	shreyasr	ANSYS	0	January 28, 2013 07:21

August 29, 2018, 08:17		#3
hand90 Member Thomas Join Date: Nov 2017 Posts: 37 Rep Power: 9	Thank you for your quick responds! I have been using monitor points and then post process the data in excel and matlab. However monitor points only gives me a snapshot of what is happening at a certain point in the domain while I am trying to see how the pressure fluctuation propagates through the entire domain (2D domain). Currently I am considering looking into ParaView but was hoping that there is a solution for this in CFD Post. Any idea?

August 29, 2018, 11:39		#6
hand90 Member Thomas Join Date: Nov 2017 Posts: 37 Rep Power: 9	I can confirm that ghorrocks method worked. 1. Choose the correct time step 2. Create a new variable (Method: Frozen Copy) 3. The new variable can then be used in any expression at any time step, keeping the "Frozen" value. So far I have not tried the second method suggested but I may try it in the future. Thank you for your help