Cumulative average

Nikpap · November 5, 2020, 12:02

Hello,

I using the implicit unsteady method to model the flow field of a burner (including the swirler).

I get 3 different velocity profiles that correspond to 3 different timesteps (10^-3, 5x10^-5 and 10^-4). It has been suggested to me to do cumulative average and see if it reaches an asymptotic value.

In star I have only seen the regular and the moving average, do you know how I could do a cumulative one?

fluid23 · November 5, 2020, 14:44

I don't think that makes sense to do inside of star-ccm. You have to do this across simulations, so its not a field function or report that any one model can return.

I had never even heard of a cumulative average outside of a GPA so I did some reading, I think what is suggested is to do this...

Y2(1) = Y(1)
Y2(2) = (Y(1) + Y(2))/2
Y2(3) = (Y(1) + Y(2) + Y(3))/3

https://www.itl.nist.gov/div898/soft...hat%20element.

Nikpap · November 6, 2020, 07:16

First off, thanks for the prompt reply.

In hindsight my original question was a bit unclear, I do apologise.

Because these 3 timesteps produce different velocity fields, I would like to see how the cumulative average evolves over the course of the simulation, say for the 10^-3 s case.

Could I do that by using a report and plot from star-ccm?

fluid23 · November 6, 2020, 09:18

That makes more sense, thanks for clarifying. I think you will probably want to check out solution histories. You can use these to store data from previous time steps. It shouldn't be too complicated of an exercise to create a field function to do a cumulative average.

Having not tried it myself, all I can do is point you in the right direction.

Check out the help files under Post-Processing>Analysis Workflow>Transient Solution Data>Transient Solution Data Workflow>Retrieving Transient Solution Data.

Perhaps someone else can give you a more specific answer.

Nikpap · February 2, 2021, 06:16

Hello!

So the way I finally did it was by creating the points I wanted to monitor in my domain. Then I created an XYZ internal table and auto extracted it based on timestep.

Using solution histories maybe could have been an option but I didnt manage to make it work and I also guess it would be way more demanding. The only downside in what I used is that I get a lot of .csv files that I have to combine in one place.

Thanks for the help!

November 5, 2020, 12:02	Cumulative average	#1
Nikpap New Member Nikolaos Papafilippou Join Date: Apr 2019 Posts: 25 Rep Power: 7	Hello, I using the implicit unsteady method to model the flow field of a burner (including the swirler). I get 3 different velocity profiles that correspond to 3 different timesteps (10^-3, 5x10^-5 and 10^-4). It has been suggested to me to do cumulative average and see if it reaches an asymptotic value. In star I have only seen the regular and the moving average, do you know how I could do a cumulative one?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
pimpleDyMFoam computation randomly stops	babapeti	OpenFOAM Running, Solving & CFD	5	January 24, 2018 06:28
Near wall treatment in k-omega SST	Arnoldinho	OpenFOAM Running, Solving & CFD	38	March 8, 2017 14:48
How to write k and epsilon before the abnormal end	xiuying	OpenFOAM Running, Solving & CFD	8	August 27, 2013 16:33
Upgraded from Karmic Koala 9.10 to Lucid Lynx10.04.3	bookie56	OpenFOAM Installation	8	August 13, 2011 05:03
Could anybody help me see this error and give help	liugx212	OpenFOAM Running, Solving & CFD	3	January 4, 2006 19:07

November 5, 2020, 14:44		#2
fluid23 Senior Member Matt Join Date: Aug 2014 Posts: 947 Rep Power: 18	I don't think that makes sense to do inside of star-ccm. You have to do this across simulations, so its not a field function or report that any one model can return. I had never even heard of a cumulative average outside of a GPA so I did some reading, I think what is suggested is to do this... Y2(1) = Y(1) Y2(2) = (Y(1) + Y(2))/2 Y2(3) = (Y(1) + Y(2) + Y(3))/3 https://www.itl.nist.gov/div898/soft...hat%20element.

November 6, 2020, 07:16		#3
Nikpap New Member Nikolaos Papafilippou Join Date: Apr 2019 Posts: 25 Rep Power: 7	First off, thanks for the prompt reply. In hindsight my original question was a bit unclear, I do apologise. Because these 3 timesteps produce different velocity fields, I would like to see how the cumulative average evolves over the course of the simulation, say for the 10^-3 s case. Could I do that by using a report and plot from star-ccm?

November 6, 2020, 09:18		#4
fluid23 Senior Member Matt Join Date: Aug 2014 Posts: 947 Rep Power: 18	That makes more sense, thanks for clarifying. I think you will probably want to check out solution histories. You can use these to store data from previous time steps. It shouldn't be too complicated of an exercise to create a field function to do a cumulative average. Having not tried it myself, all I can do is point you in the right direction. Check out the help files under Post-Processing>Analysis Workflow>Transient Solution Data>Transient Solution Data Workflow>Retrieving Transient Solution Data. Perhaps someone else can give you a more specific answer.

February 2, 2021, 06:16		#5
Nikpap New Member Nikolaos Papafilippou Join Date: Apr 2019 Posts: 25 Rep Power: 7	Hello! So the way I finally did it was by creating the points I wanted to monitor in my domain. Then I created an XYZ internal table and auto extracted it based on timestep. Using solution histories maybe could have been an option but I didnt manage to make it work and I also guess it would be way more demanding. The only downside in what I used is that I get a lot of .csv files that I have to combine in one place. Thanks for the help!