What do you CFD guys do during a long simulation running?
 

A stupid but common question.

From my knowlewdge, CFD code running is a very special job compared with many other subjects. FEA code often takes seconds to finish, and most experimental work can go instantly because it's done by the God. But CFD code running usually costs hours, days or even more. People have to spend time waiting before they can begin the next step.

Just wonder what do you CFD guys do during the long computer running?

First, of course we can relax, take a break or go to restroom. But I think people have to be attendant more or less, checking progress, if it's diverging or if the tendency is correct. And we may stop the running periodically, backup and postprocess the interrim data a little bit.

Second, reading books or papers during this time sounds not bad. But I feel like it's not easy concentrating on reading when you've to care for the computer also.

Thrid, I've bad experience if I went home for sleep or went for a vacation during this boring time. May be I was just in bad luck. The simulations often went wrong, diverged, or computer got into unexpected trouble (stall or restarted by others).

Therefore, running a CFD code is a painful work. Any good advice?

i do turbulent direct numerical simulation(DNS), it takes me about one month to get a result. During this period, i have many thing to do, reading paper, write code is my main job.
LOL.

What? No, don't break the illusion that we are actually working and not just simulating work as well! ;)

Play Civ IV, watch Brit shows like Screen Wipe, 8 out 10 Cats etc.

Postprocess for earlier runs..............

pr0n
 

That's the best way to pass the time! ;)