CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > OpenFOAM > OpenFOAM Post-Processing

Different results using sample utility vs Paraview 'Plot over line' tool

Register Blogs Community New Posts Updated Threads Search

Like Tree1Likes
  • 1 Post By thomasArk47

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   March 18, 2016, 09:15
Exclamation Different results using sample utility vs Paraview 'Plot over line' tool
  #1
New Member
 
pulkit
Join Date: Jun 2013
Posts: 5
Rep Power: 13
agarwa58 is on a distinguished road
Hi guys,

I am trying to get the accurate values of air velocities at specific points in the flowfield for a flow through a contracting duct simulation using simpleFoam. In order to compare the results against a commerical solver, I am evaluating teh velocities only at cell centres without using interpolation. However, the answers obtained by using :
1. Sample utility (with 'cell' interpolation option)
2. Paraview plot over line tool (using only cell data)

are quite different from one another. Qualitatively the flwo field looks accurate enough but my validation study is highly dependent on the result of these point velocities.

Can anyone tell which one of the two options (sample utility vs paraview) is more accurate in evaluating point velocities in the flow field ??

Thanks,
Pulkit
agarwa58 is offline   Reply With Quote

Old   March 18, 2016, 13:29
Default
  #2
Member
 
Thomas Boucheres
Join Date: May 2013
Posts: 41
Rep Power: 13
thomasArk47 is on a distinguished road
Hi,

have you tested the results using other interpolation procedures in OpenFoam (cellPoint, cellFacePoint)? It can be a highly critical parameter if the mesh is "coarse" in view of the fluctuations of the quantity you want to sample.

On the other side, I think the better is to extract your required lines from OpenFoam not from Paraview since there is more "control" on the accuracy side of algorithmic procedures in OpenFoam than on Paraview.
thomasArk47 is offline   Reply With Quote

Old   March 19, 2016, 11:32
Default
  #3
Senior Member
 
Join Date: Jun 2012
Location: Germany, Bochum
Posts: 230
Rep Power: 16
Bazinga is on a distinguished road
if you take the sample utility with cell as an interpolation scheme you will get the value which is calcuated and stored in the nodal point in the middle of the cells for each point that is inside this cell. so with this option there is no interpolation used. it is the "truest" result you can present.

Last edited by Bazinga; March 20, 2016 at 08:13.
Bazinga is offline   Reply With Quote

Old   March 19, 2016, 13:44
Default
  #4
Member
 
Thomas Boucheres
Join Date: May 2013
Posts: 41
Rep Power: 13
thomasArk47 is on a distinguished road
Hello Bazinga,

my english being poor, perhaps i do a mistake but I'm not agree with your "truest" notion.

Indeed, taking "cell" option in sampleDict doesn't perform interpolation but from accuracy point of view it is clearly the "worst" case possible, not the "truest".
Whatever the precise location of your target sample point is in a cell, the sampled value will be the same since we are taking the cell value. It will be clearly better when using stencil (points, faces) to build a more accurate interpolation. Morever, dealing with a finite volume basis, don't forget that the cell value cannot be better than a mean value in the cell.
Bazinga likes this.
thomasArk47 is offline   Reply With Quote

Old   March 20, 2016, 08:24
Default
  #5
Senior Member
 
Join Date: Jun 2012
Location: Germany, Bochum
Posts: 230
Rep Power: 16
Bazinga is on a distinguished road
Dear Thomas,

You are right with most points you made but let me clarify two things:
By "truest" I meant that using interpolation schemes is already post processing the data and it shows manipulated results of the FVM. The word "truest" is probably not a good term here.

In my (limited) experience I found that using interpolation schemes in the sample utility does not necessarily lead to the best results. For example it happened to me that crucial information got lost due to interpolation where a small recirculation area above a building was not shown when using interpolation while the cell values clearly showed negative values. In this case the cell values were clearly not the worst results.

In any case if you present the data using interpolation name the scheme that was used.
Bazinga is offline   Reply With Quote

Old   March 20, 2016, 09:57
Default
  #6
Member
 
Thomas Boucheres
Join Date: May 2013
Posts: 41
Rep Power: 13
thomasArk47 is on a distinguished road
Hello Bazinga,

your example of a negative value not seen by interpolated sampled values is good and interesting. Indeed, using direct cell scheme for sampling can be usefull for some "sanity check" (something like debugging a solver...) since in this case the sampling gives directly the variables as they are used during the finite volume integration. The "truest" ones indeed. From this point of view, i now understand the meaning of your expression.

On the other side, for a benckmark study (using a validated algorithm which is known to be convergent say), it is better to use high order sampling (using the same for all comparative cases)

Maybe agarwa58 will give us more informations on his problem (if it is yet one)
thomasArk47 is offline   Reply With Quote

Reply

Tags
paraview, sample utility


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 Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
[Gmsh] Problem with Gmsh nishant_hull OpenFOAM Meshing & Mesh Conversion 23 August 5, 2015 03:09
'sample' utility for 'U' yields different results for simple-scotch-etc. HakikiCanakkaleli OpenFOAM Post-Processing 3 January 5, 2014 13:08
[OpenFOAM] Annoying issue of automatic "Rescale to Data Range " with paraFoam/paraview 3.12 keepfit ParaView 60 September 18, 2013 04:23
[blockMesh] Axisymmetrical mesh Rasmus Gjesing (Gjesing) OpenFOAM Meshing & Mesh Conversion 10 April 2, 2007 15:00
error while compiling the USER Sub routine CFD user CFX 3 November 25, 2002 16:16


All times are GMT -4. The time now is 12:40.