Space-time post processing
 

Hey,

How would you present data from a simulation where the time-step is not constant in space? Is it even possible to have any meaningful information other than the steady-state solution?

Hi,

Have you considered to use average values over a certain time interval?

Best regards,
Marcus

Hej Marcus, thank you for your input. The answer is no I have not. The problem is that time is progressing faster in some parts of the domain. Normally this is solved by some kind of book-keeping procedure, performing several time-steps in the regions where time progresses slower. With that kind of procedure the global time would be the same and the results would be more tangible (at least for me ;)). I don't see how time-averaging would help unless I use the book-keeping procedure. The time-average values would not be of the same global time, but perhaps that is not a problem?

Hi,

I thought that you could take the time average for the results for lets say 0-5 s global time over the whole domain. The only difference would then be that you have more solutions to take average of in the domain that has smaller time steps and fewer in that with longer time steps. If this does not work I guess the book-keeping procedure is your best option.

Best regards,
Marcus