tecplot output + ParFEAP

[simapise]

Dear all,

parFEAP writes some elements twice in the output data for Tecplot. I guess those are ghost elements for each zone/partition. It looks very bad in Tecplot because it results to dark boundaries (overlapped elements at zone boundary) of each zone/partition. Does anybody have an idea how could i solve this problem? How could i avoid writing ghost or double elements in output file? If it is not possible then which is the best way for post processing of simulation data from parFEAP?

Thanks in advance.

[Prof. S. Govindjee]

I use paraview and the PVIEw command for writing the *.vtu files.  Have you tried that and encountered 'errors'?

[simapise]

now I managed get parFEAP out for Paraview. It works perfect for serial FEAP using PVIE command but the command PVIE,time doesn't work.

Then I tried it for parFEAP using PVIE command but the command PVIE,time doesn't work there also. parFEAP using PVIE writes one .vtu files for each time step but it is not written well. NumberOfPoints and NumberOfCells are not correct and much less than it should be. When i try to open pv.pvd file in paraview, it gives lot of errors.

The batch file i use is as follows

nopr
usol,pard,1,0,6
PVIE
auto,time,4,7,500
dt,,1e-4
auto,dt,1e-10,1e-0
loop,,10000
  loop,1
   time
    loop,,10
    tang,,1
    next
    save
plot,hide
plot,stre,24,,1
plot,hide
next
PVIE
next

Please guide me to go further.

[Prof. S. Govindjee]

What version of FEAP are you using?

[simapise]

thanks for quick reply, I am using FEAP v82.

[Prof. S. Govindjee]

8.2 is very old.  I'm not sure if pview will function properly with it.  You should probably upgrade your code to the newest version: http://feap.berkeley.edu/wiki/index.php/Obtaining_FEAP

[blackbird]

Hi.

I know this topic is quite old, however I spent some time on the issue FEAP --> Tecplot and I would like to share my knowledge:

The overlapping element  issue is quite annoying as it looks awfull. afaik there are two ways to deal with it:

1. separate your tecplot into mesh and solution. The mesh can be written during the creation of the partitions for parfeap, where there is the whole mesh in a single feap-process.

2. If you require element-based data for your solution-files, then a global element-ID may be the choice. Again, during the partitioning, I created a special file for each partition, that includes the local to global element-ID. Whenever the tecplot solution is output, read that file to get the global-ID and sort the information of that element into the element information array

Feel free to ask if more detailed information on both approaches is required