|
[Sponsors] |
![]() |
![]() |
#1 |
Guest
Posts: n/a
|
i want to show variables in the center of mesh,but the result showed by TECPLOT is on node,why? the code is below:
write(33,*) 'TITLE = "3D"' write(33,*) 'VARIABLES = "X" "Y" "Z" "Rho" "U" "V" "W"' write(33,*) 'DATAPACKING=BLOCK, ZONETYPE=FEBRICK,N=',numnp,',','E=',nelem,',','VAR LOCATION=([4-7]=CELLCENTERED)' ! WRITE(33,*) 'DT=(DOUBLE DOUBLE DOUBLE DOUBLE DOUBLE DOUBLE DOUBLE)' !----------------------BLOCK TYPE--------------------------------- DO I=1,NUMNP WRITE(33,*) CX(1,I) END DO DO I=1,NUMNP WRITE(33,*) CX(2,I) END DO DO I=1,NUMNP WRITE(33,*) CX(3,I) END DO DO I=1,NELEM WRITE(33,*) uh(1,I,0.D0,0.D0,0.D0) END DO write(33,'(///)') DO I=1,NELEM WRITE(33,*) uh(2,I,0.D0,0.D0,0.D0)/uh(1,I,0.D0,0.D0,0.D0) END DO write(33,'(//)') DO I=1,NELEM WRITE(33,*) uh(3,I,0.D0,0.D0,0.D0)/uh(1,I,0.D0,0.D0,0.D0) END DO write(33,'(//)') DO I=1,NELEM WRITE(33,*) uh(4,I,0.D0,0.D0,0.D0)/uh(1,I,0.D0,0.D0,0.D0) END DO !----------------------------------------------------------------- write(33,'(/)') do i=1,nelem write(33,'(<ncount(i)>i10)') nname(i,1),nname(i,3),nname(i,7),nname(i,5),nname( i,2),nname(i,4),nname(i,8),nname(i,6) end do close(33) |
|
![]() |
![]() |
![]() |
#2 |
Guest
Posts: n/a
|
i encountered that problem too. while i had given the variable locations as cellcentered, it interpolated the values to boundary nodes. well, i doesn't matter most of the time but at the boundary for instance it extrapolates from cell centers to cell boundaries, consequently it seems that the result is wrong for example in case of a fixed value boundary condition. using ghost cells and writing the results including the ghost cells will resolve the problem but this is just a specific solution, not a general purpose one.
|
|
![]() |
![]() |
![]() |
#3 |
Guest
Posts: n/a
|
i do not work with finite element option. for structured grid by control volume method: 1-load data to tecplot 2-select the zone(s) 3-"points" tab=>"points to plot"=>select "all cell centers"
|
|
![]() |
![]() |
![]() |
#4 |
Guest
Posts: n/a
|
u should understand that, vertex data is requirement of visualization procedure, in which we need we have some shape function which interpolate color value.
considering this, a cell-centered data is equivalent to picewise-constant interpolation of colors, i.e., we have a fixed color in each cell and so do usually not have a desirable plot. So you should provide a vertex data (do interpolation of your cell centered data to vertexes) in a consistent manner, if plotter do this for you, it would not be essentially desirable. btw, Paraview (VTK) (is free) has option for cell-centered data (as i said, picewise constant interpolation), PS: i use both of paraview and tecplt, paraview is indeed powerful, but bit more difficult to use in particular for begineer PS: for paraview file format see this: http://www.vtk.org/pdf/file-formats.pdf |
|
![]() |
![]() |
|
|
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
FvMatrix coefficients | shrina | OpenFOAM Running, Solving & CFD | 10 | October 3, 2013 14:38 |
gradient calculation of cell centered finite volume method | zhengjg | Main CFD Forum | 10 | November 11, 2012 23:13 |
Cells with t below lower limit | Purushothama | Siemens | 2 | May 31, 2010 21:58 |
[Commercial meshers] Trimmed cell and embedded refinement mesh conversion issues | michele | OpenFOAM Meshing & Mesh Conversion | 2 | July 15, 2005 04:15 |
Warning 097- | AB | Siemens | 6 | November 15, 2004 04:41 |