Hi! All.

I am working on the simulation of buoyancy convection under microgravity. In our case, the gravity is a function of time (in the simplest case, a cos or sin function). We are using Finite Volume Method to study the case. Our problem is that the time comsumption is unaccetpable. The reason is our time step will be restricted by the frequency of gravity variation. For example, when the frequency is 1/8 HZ, the time step has to be less than 1 second inder to track the gravity variation. And in order to obtain a steady, about 100000 time steps are required. It really cost a lot of time.

Has anyone met similar problems? Is there any better algorithm recommendation? Your help will be highly appreciated.

If your method has CFL 1 restriction, then probably your time step is 8/100000 such that you need 100000 steps to finish a period. 100000 is often seen in transitional flow studies. Parallel computation is indispensible now.

Thank you very much. Is parallel computation the only way i can choose? How about spectral anaylsis?