CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > OpenFOAM Running, Solving & CFD

uniform to nonuniform interpolation

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   April 9, 2010, 10:45
Default uniform to nonuniform interpolation
  #1
Member
 
Kim Yusik
Join Date: Dec 2009
Posts: 39
Rep Power: 7
impecca is on a distinguished road
Dear Foamers.

I want to interpolate data ('origin' data, which are placed on a UNIFORM mesh) into a non-uniform patch('copied' data) - e.g. generate turbulent inlet velocity on patch(array) which I generate inside of the code and interpolate to real(non-uniform) geometry.

I checked several interpolation utilities in OF (/src/OpenFOAM/interpolations) and I think the most closest one is patchToPatchInterpolation.

Based on Jasak's post(Channel flow) this utility take 'origin' patch from the geometry. but I need to make uniform mesh which is not part of the geometry.

Is there anyone can give me an advice How to interpolated data from unifrom mesh(user generated, which is not part of geometry) to non-uniform mesh ?

thank you very much.

Yusik
impecca is offline   Reply With Quote

Old   April 14, 2010, 06:26
Default
  #2
Senior Member
 
Ivan Flaminio Cozza
Join Date: Mar 2009
Location: Torino, Piemonte, Italia
Posts: 207
Rep Power: 9
ivan_cozza is on a distinguished road
Send a message via MSN to ivan_cozza
Quote:
Originally Posted by impecca View Post
Dear Foamers.

I want to interpolate data ('origin' data, which are placed on a UNIFORM mesh) into a non-uniform patch('copied' data) - e.g. generate turbulent inlet velocity on patch(array) which I generate inside of the code and interpolate to real(non-uniform) geometry.

I checked several interpolation utilities in OF (/src/OpenFOAM/interpolations) and I think the most closest one is patchToPatchInterpolation.

Based on Jasak's post(Channel flow) this utility take 'origin' patch from the geometry. but I need to make uniform mesh which is not part of the geometry.

Is there anyone can give me an advice How to interpolated data from unifrom mesh(user generated, which is not part of geometry) to non-uniform mesh ?

thank you very much.

Yusik
Dear Yusik,
do you success to interpolate your stuff? It works also in parallel simulations? Btw, why do you need to generate your turbulent inlet on a uniform mesh? I'm working on a similar argument, and I'm trying to generate the inlet on a general patch. There are ways (like the Billson's approach to synthetic turbulence, or even with digital filters, but only on scalar simulations) to do that on general patches.

Regards, Ivan
ivan_cozza is offline   Reply With Quote

Old   April 15, 2010, 12:43
Default
  #3
Member
 
Kim Yusik
Join Date: Dec 2009
Posts: 39
Rep Power: 7
impecca is on a distinguished road
Quote:
Originally Posted by ivan_cozza View Post
Dear Yusik,
do you success to interpolate your stuff? It works also in parallel simulations? Btw, why do you need to generate your turbulent inlet on a uniform mesh? I'm working on a similar argument, and I'm trying to generate the inlet on a general patch. There are ways (like the Billson's approach to synthetic turbulence, or even with digital filters, but only on scalar simulations) to do that on general patches.

Regards, Ivan

Dear Ivan

answering first.
- Do I success ?: No, I am still struggling to find the way to mapping turbulence data which I made (I use my supervisor method which is based on Klein's method i.e. disital filters)
- Working on parallel ? :excuse my ingorance. I have no idea how's the generating turbulence cause a problem in parallel computing. I also read your post (Synthetic turbulence BC for LES) but still I don't get it, may be it because which source code we are based on. I modifed pisoFoam.C which means I directly manipulated patch field inside of solver code. I tried to make new BC for this but I couldn't and I am still working on it.
- why do I need uniform mesh? :Basically I am looking for which is less painful work for my case. So I thought that if I can mapping generated turbulence data into real geometry (channel, structure and nonuniform) that will be the easiest way. becasue my filters(I hope you would know the meaning of filters) and random number arrays in the code are generated based on uniform grid.
may be if you can let me know how can I generate filters and random filed and how can I manipulate them with integral length scale on general patch then I don't need to uniform-nonuniform mapping tool. (I know this is very,... general and silly question so if you let me have a chance specify this question then I'll stess about the detail problem which I had)
and also if inlet is uniform mesh, then also I donot need a mapping.

now my questions.
- There are ways (...) to do that on general patches. : do you know is there any open source for turbulence generator which use digital filter methods. and could you tell me what is the meaning of scalar simulation? does it mean it is impossible for velocity field?


plus
- I know one site which contains turbulence generator code (I think this is based on klein's method. http://www.ltt-rostock.de/mediawiki/...tockExtensions) you can download source file from there.

- I know one phD student who is working on wind turbine acoustics using synthetic turbulence though he doen't use openfoam. but if you want to discuss something with him, I can ask him.

Thanks
impecca is offline   Reply With Quote

Old   April 16, 2010, 03:33
Default
  #4
Senior Member
 
Ivan Flaminio Cozza
Join Date: Mar 2009
Location: Torino, Piemonte, Italia
Posts: 207
Rep Power: 9
ivan_cozza is on a distinguished road
Send a message via MSN to ivan_cozza
Dear Kim,
I'm workin on a method that's derived from the Klein method too, so nice to know that there are other guys struggling on it!
First, some answers:

- Yes, filters can do it bad in parallel computing if you implement it without doing nothing! If your inlet patch is shared from more than one processor, when you are close to the processorBoundary yo have to look in faces that are in different processors, and the parallel structure of foam is not ready for that, as it memorize only the data close to the boundary.

- I wrote down a solver for my synthetic turbulence from zero, as Im interested in generating turbulence inside the field, not only at the patches (of course, I also created a patch for inlet turbulence from inside the solver... I'm really a bad c++ programmer, I hate to change it in a real bc, but I know it would be better...).

- Yes, you can do filters also on general patches. Up to now I do it resolving the full integral inside the convolution, without exploiting the separability of the gaussian filter (now, it's bad for computational cost, but you can do on whatever patch you want). I'm looking for better solutions, one would be do it like you, do it on a structured mesh, generated from inside the code, and then map the result on the general mesh, but another time you have to spend time for mapping, and parallel computing is hard...

- Scalar computation -> not parallel!

Now some questions:

- Just for chnowledge, who is your university and supervisor?

- If you can, I would like to talk with your friend that works on wind turbines, I'm involved in a similar problem (acoustics of turbomachinery airfoils), so communications is always good!

- Thanks for the link, I'll take a look on it soon!

Have a nice day, Ivan



Quote:
Originally Posted by impecca View Post
Dear Ivan

answering first.
- Do I success ?: No, I am still struggling to find the way to mapping turbulence data which I made (I use my supervisor method which is based on Klein's method i.e. disital filters)
- Working on parallel ? :excuse my ingorance. I have no idea how's the generating turbulence cause a problem in parallel computing. I also read your post (Synthetic turbulence BC for LES) but still I don't get it, may be it because which source code we are based on. I modifed pisoFoam.C which means I directly manipulated patch field inside of solver code. I tried to make new BC for this but I couldn't and I am still working on it.
- why do I need uniform mesh? :Basically I am looking for which is less painful work for my case. So I thought that if I can mapping generated turbulence data into real geometry (channel, structure and nonuniform) that will be the easiest way. becasue my filters(I hope you would know the meaning of filters) and random number arrays in the code are generated based on uniform grid.
may be if you can let me know how can I generate filters and random filed and how can I manipulate them with integral length scale on general patch then I don't need to uniform-nonuniform mapping tool. (I know this is very,... general and silly question so if you let me have a chance specify this question then I'll stess about the detail problem which I had)
and also if inlet is uniform mesh, then also I donot need a mapping.

now my questions.
- There are ways (...) to do that on general patches. : do you know is there any open source for turbulence generator which use digital filter methods. and could you tell me what is the meaning of scalar simulation? does it mean it is impossible for velocity field?


plus
- I know one site which contains turbulence generator code (I think this is based on klein's method. http://www.ltt-rostock.de/mediawiki/...tockExtensions) you can download source file from there.

- I know one phD student who is working on wind turbine acoustics using synthetic turbulence though he doen't use openfoam. but if you want to discuss something with him, I can ask him.

Thanks
ivan_cozza is offline   Reply With Quote

Old   April 21, 2010, 05:24
Default
  #5
Member
 
Kim Yusik
Join Date: Dec 2009
Posts: 39
Rep Power: 7
impecca is on a distinguished road
Thanks for the detail.

May be I could meet same problem with you in the near future. I'll keep eye on parallel computing.

my supervior is Dr. Xie, Zheng-tong in U of Southampton. UK
and I also send email to my colleague and he my contact you through this forum.

lastly, about the mapping, I just mapping without using any interplation options in OF. I just extract y & z positions from the patch(non uniform) and make them track turbulence data (uniform) which lies equivalent position. (using traditional for loop and if statement.)

thank you again and keep in touch.

Yusik
impecca is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
boundary conditions for simpleFoam calculation foam_noob OpenFOAM Running, Solving & CFD 8 July 1, 2015 08:07
Pressure instability with rhoSimpleFoam daniel_mills OpenFOAM Running, Solving & CFD 44 February 17, 2011 18:08
rhoSimpleFoam claco OpenFOAM 7 April 20, 2010 04:32
FOAM FATAL IO ERRORsimpleFoam hariya03 OpenFOAM Running, Solving & CFD 6 July 16, 2008 08:03
RasInterFoam STRANGE RESULTS AT BOUNDARY kumar2 OpenFOAM Running, Solving & CFD 8 March 24, 2008 19:38


All times are GMT -4. The time now is 17:40.