|
[Sponsors] |
Frequency analysis of a Fluidic Oscillator at output |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
May 21, 2016, 11:57 |
Frequency analysis of a Fluidic Oscillator at output
|
#1 |
New Member
Daniel
Join Date: May 2016
Posts: 5
Rep Power: 10 |
Hello everyone,
Situation: I'm simulating a Fluidic oscilaltor (see image in attachment taken from google) with pimpleFoam. For pre-processing, I used "icem cfd" to make the geometry and mesh, export it as a fluent-file and then imported it to openFoam, this was no problem. The most important property I need to analize is the frequency of the fluid-oscillation at the output. For that, I thought I could plot de integration of the velocity field over the time on "output" using patchIntegrate. Problem: I defined the two outlets of the oscillator (right side of image) as one patch named "output" in icem. So when I use patchIntegrate, it integrates the velocity over both outlets and therefore it's not clear to analize the frequency of oscilaltion Posibilities?: I thought of 3, though I'm not sure if they can be implemented. If you think of more, they are very welcome! 1. Define a region when using patchIntegrate. Does the patch has to be devided in regions beforehand? How to define the "regions"? 2. Use a tool from paraview to plot the massflow through a part of the "outlet"-patch. How is the tool called? 3. Extract the cell velocity-data from the time-directories of the desired region "outlet" and make the integration mannually. (no clue how to do that) Also, in case you have another method to analize the frequency, they are very welcome as well! Thanks a lot in advance for your help Daniel |
|
May 23, 2016, 05:11 |
|
#2 |
Senior Member
|
For method 2:
First make a clip that contains only 1 half, after that use the filter Integrate Variables and finally save data from all time steps in .csv files. Usually you end up with 1 csv file for each time step, so you should make some script to combine them into one file. Then use some Fast Fourier Transform library. I think python should work for both the combination into one file and the FFT. Regards, Tom |
|
May 23, 2016, 06:11 |
|
#3 |
Senior Member
Joachim Herb
Join Date: Sep 2010
Posts: 650
Rep Power: 22 |
I think you can use createPatch (https://openfoamwiki.net/index.php/CreatePatch and look for it in this forum) to create two new boundaries (e.g. outlet1, outlet2) from the existing one (actually, why didn't you do this from the beginning?). Then you can use patchIntegrate on each of them. If you have already done your simulation, you might be able to use
execFlowFunctionObjects -dict system/functionObjectsDict -noFlow to just do the postprocessing (but I am not sure if this works with a modified mesh). |
|
May 23, 2016, 06:41 |
|
#4 | |
New Member
Daniel
Join Date: May 2016
Posts: 5
Rep Power: 10 |
Quote:
Do you mean first make a clip of the one half and after that a slice at the output, and then to that slide apply the "Integrate variables"?... otherwise I wouldnīt know how to do it just with the clip. I have around 520 time steps. so that means I would cahnge manually every time step and then export each result form "Integrate variables" to a csv ? Thanks a lot, as you notice, Iīm new in CFD |
||
May 23, 2016, 06:45 |
|
#5 |
Senior Member
|
First only open your outlet patch (Or use the extract block filter to have only the outlet patch) then use the clip on that.
|
|
May 23, 2016, 07:01 |
|
#6 | |
New Member
Daniel
Join Date: May 2016
Posts: 5
Rep Power: 10 |
Quote:
You are right, I should have done it from the beginning. As the simulation needs a lot of time to complete, Iīm trying to find another solution now I checked the website, I seems like a very good alternative. Do you know if thereīs a tutorial example that uses this utility? I didnīt checked how to modify it specifically for my case, as it seems the have some sort of cylindrical patch in the script-example Thanks! |
||
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
3D analysis of Ahmed body | Irshad22 | FLUENT | 0 | December 17, 2009 05:33 |
Frequency analysis | Usman | CFX | 0 | March 16, 2008 19:36 |
Questions on Analysis Controls and Runtime Control | Arnab | Siemens | 1 | October 26, 2004 10:47 |
Short Course: Computational Thermal Analysis | Dean S. Schrage | Main CFD Forum | 11 | September 27, 2000 18:46 |
Is CFD Science or Art ? | John C. Chien | Main CFD Forum | 36 | October 5, 1999 13:58 |