Average Cd through transient

smartsoldier · July 21, 2014, 09:59

When I run my transient solution I have automatic export of data every time-step and statistics on in the iterate panel. I also configured the first plot window to show the Cd at every time-step. How can I obtain these Cd values? Or better yet obtain the average of all of them over my time-steps automatically?

Edit: If I load my exported solution file into CFD-Post I can go to the calculators tab and run the function calculator for force on the object I get what I want. But I would have to manually load and process each calculation by hand, which is not possible when my solution has 100's of timesteps.

flotus1 · July 21, 2014, 10:10

Quote:

Originally Posted by smartsoldier

I also configured the first plot window to show the Cd at every time-step.

In the same menu tick the box "write to file".
You get a text file with the raw data that can be processed any way you like.
There is also an option to automatically average monitor values over a specified amount of samples, at least in the most recent versions of Fluent.

smartsoldier · July 21, 2014, 12:51

Thanks for the help! The raw data has multiples of each time step and different values of drag for each one. Why is that? Also, could you point me to the "option to automatically average monitor values over a specified amount of samples"?

Edit: Upon closer inspection, it seems the first half or so raw values are identical between the different design points, only the second half change. Why is that?

flotus1 · July 21, 2014, 15:50

The data file fluent creates is never overwritten.
New data is simply appended at the end of the existing file.

Quote:

Also, could you point me to the "option to automatically average monitor values over a specified amount of samples"?

I thought this option could be found in the monitors setup window aswell.
I can check again tomorrow.

smartsoldier · July 21, 2014, 16:52

The data file is rewritten for every design point I think. I have a main project file that runs the simulation for 5 other design points. This makes 5 separate folders and in each is a new txt file containing the drag information. As I said some of the raw data is repeated, for example 5 different cd for a single value of timestep. Some of the data files are longer than others. Any ideas why?

flotus1 · July 22, 2014, 03:39

The option "Average Over(Time Steps)" is where I suspected it to be, in the monitor setup window.

I can only guess what causes fluent to write several different values for the same time step. Maybe you somehow managed to export the values for each inner iteration. I dont know how this could be achieved, but it would explain why the data files dont have the same length (provided they should contain the same amount of time steps).

Wait a minute...
Did you set up the drag monitor using the "Residuals, Statistics and Force Monitors" or did you use the "Surface Monitors" instead? If you used the latter, this could explain what happened. Here you can indeed choose to get data every iteration but to use the time steps as x-axis at the same time.

cfmyd · July 22, 2014, 17:46

Write your drag values to a file, like flotus1 said. From there you can easily set up functions in a language of your choice to read in the drag values out of that file. I use MATLAB's "importData", and it loads the data from that file into a matrix. From there it is easy to get the average over as many time steps as I want.

July 21, 2014, 09:59	Average Cd through transient	#1
smartsoldier New Member Jacob Join Date: Apr 2014 Posts: 23 Rep Power: 12	When I run my transient solution I have automatic export of data every time-step and statistics on in the iterate panel. I also configured the first plot window to show the Cd at every time-step. How can I obtain these Cd values? Or better yet obtain the average of all of them over my time-steps automatically? Edit: If I load my exported solution file into CFD-Post I can go to the calculators tab and run the function calculator for force on the object I get what I want. But I would have to manually load and process each calculation by hand, which is not possible when my solution has 100's of timesteps.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Time average results in Transient CFX Simulation	BalanceChen	CFX	32	September 30, 2021 13:59
Time average Result in transient Run	omidiut	CFX	7	July 28, 2013 07:35
Average value of expression in transient flow analysis	JFK	CFX	1	May 22, 2013 08:52
Average of Expressions in Transient Run for Parameter Study	BigPapi34	CFX	1	August 7, 2012 05:34
How obtain the average of lift over time for a transient simulation?	aero	ANSYS	0	November 11, 2009 02:00

July 21, 2014, 12:51		#3
smartsoldier New Member Jacob Join Date: Apr 2014 Posts: 23 Rep Power: 12	Thanks for the help! The raw data has multiples of each time step and different values of drag for each one. Why is that? Also, could you point me to the "option to automatically average monitor values over a specified amount of samples"? Edit: Upon closer inspection, it seems the first half or so raw values are identical between the different design points, only the second half change. Why is that?

July 21, 2014, 16:52		#5
smartsoldier New Member Jacob Join Date: Apr 2014 Posts: 23 Rep Power: 12	The data file is rewritten for every design point I think. I have a main project file that runs the simulation for 5 other design points. This makes 5 separate folders and in each is a new txt file containing the drag information. As I said some of the raw data is repeated, for example 5 different cd for a single value of timestep. Some of the data files are longer than others. Any ideas why?

July 22, 2014, 03:39		#6
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,399 Rep Power: 46	The option "Average Over(Time Steps)" is where I suspected it to be, in the monitor setup window. I can only guess what causes fluent to write several different values for the same time step. Maybe you somehow managed to export the values for each inner iteration. I dont know how this could be achieved, but it would explain why the data files dont have the same length (provided they should contain the same amount of time steps). Wait a minute... Did you set up the drag monitor using the "Residuals, Statistics and Force Monitors" or did you use the "Surface Monitors" instead? If you used the latter, this could explain what happened. Here you can indeed choose to get data every iteration but to use the time steps as x-axis at the same time.

July 22, 2014, 17:46		#7
cfmyd New Member James Join Date: Jun 2014 Posts: 3 Rep Power: 11	Write your drag values to a file, like flotus1 said. From there you can easily set up functions in a language of your choice to read in the drag values out of that file. I use MATLAB's "importData", and it loads the data from that file into a matrix. From there it is easy to get the average over as many time steps as I want.