I am doing a run with 3M+ fluid elements, and when I launch the analysis, I get the notification that the first 25 iterations run in just a few hours. Then the next 25 take about a day, and then it takes almost a day to run just a couple iterations. So I stopped the run, and re-started it just to see what would happen. It did almost the exact same thing- The first 25 runs of the same model after the restart took several hours, the second 25 took a bit longer, and now it is crawling.
I have not seen this before on large runs. They tend to solve at a fairly uniform rate (or crash and burn somewhere along the way).
Any words on if this is normal or has a known cause?
BTW- The analysis is being run remote on a dedicated 1-socket, 4-core, 2.8GHz machine with 12GB memory running 64-bit Windows 7 OS.
Thanks for any comments.
The only cause I can think of is that you are running transient and filling the HDD up?
If not, feel free to share the support file (cfz) so we can take a look here.