Hi,

I'm using Fluent running on a Linux cluster that uses PBS to parallelize it. The problem is the data file is extremely large, and I only uses a small portion of it. Is there any way to tell Fluent, or my script file, that it only needs to save certain data, such as just pressure and velocity, instead of writing everything?

I have set the autosave function (file->write->autosave) in FLUENT by setting it up when I setup all of my other values in preparation for submitting an unsteady PBS job...and it works. If you have created a surface, line, rake or point to extract the data needed before submitting job than one should (although, I haven't tried it yet with PBS) be able to set it up using Solve->monitors->surface . It can return a value you designate every timestep (no way of telling it once per x timesteps).

I tried it (in "regular" FLUENT) with the rake function but it doesn't write values at all points on the rake but either sums or averages all the points, but if you don't have too many you could always just author your own points (surface->point) you want and do it that way.

Cheers, Tim.

hi Ross,

Since ur using a PBS script to parallelise fluent, you may have a Input File, where-in u mention to write the newly iterated values that is .cas and .dat file, here u can mention just only to write the case file alone and also use the Auto save option to write a output file with ur required parameters, r even use a monitor to monitor the values, wherein u can write it after some prescribed iterations, I have not tried this, hope this helps.

krishna.