CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ParaView (https://www.cfd-online.com/Forums/paraview/)
-   -   [OpenFOAM] Animation always starts from zero (https://www.cfd-online.com/Forums/paraview/126278-animation-always-starts-zero.html)

McFly November 13, 2013 01:27

Animation always starts from zero
 
Hello,

I have the following two issues with paraFoam running on a cluster with OF 2.1.1.

I want to save out images of each timestep for an iso-surface. However, when I run the "save animation" sequence after about 60 images it starts saving images of the background without the iso-surface. So it works fine initially, then stuffs up.

The second issue occurs when I try to restart the animation sequence from the missing images. I use the animation view to select the sequence: start time, end time, no of frames. But when I run this sequence it ALWAYS starts playing from timestep 0, with weird intervals up to the end time.

My current work around is to remove the directories it has already output correctly and then just run it again - but this is silly.

Does anyone have an answer to these issues?

Thanks,


Marty

McFly November 13, 2013 20:21

The second issue seems to have fixed itself. Logging in last night and today the animation panel works correctly.

However the first issue remains - Paraview starts saving images without surface content after about 50 timesteps. Any ideas?

No errors are reported.


Marty

startingWithCFD November 14, 2013 03:28

Hey there! I also had problems in the past with paraview animations (freezing after some frame, probably due to memory limitations). Two remedies may be to only load what you need (for example I only loaded the outside patch) and to play around with the settings (I think the "Cache geometry" setting made a big difference in my case). I hope this helps.

mturcios777 November 14, 2013 12:39

Hello Martin,

Are you sure that the iso-surface still exists at the timesteps where it dissapears from the simulation? Just trying to eliminate all possible sources of error, and I've had that happen where I haven't adjusted my range properly.

McFly November 18, 2013 19:19

Hello,

Thanks for the replies.

The surface exists for all timesteps. As I was in a hurry I ran the animation for 50 timesteps several times (only 300 timesteps total) to produce the required frames.

I will have to try the 'cache geometry' option - memory usage is probably the culprit given the size of the dataset.


All times are GMT -4. The time now is 22:33.