Expanding the physical domain #2211
-
Hello AMReX people, I am implementing far field boundary conditions using FMM for velocity-vorticity formulation. For that, I need to confine the vorticity region within my physical domain. Thus, when the vorticity reaches the physical boundary, I have to expand the physical domain in that direction on the coarsest grid to keep the vorticity confined. Using AMReX, is it possible to do that on the fly? Thank you in advance. Shahab |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 5 replies
-
I don't know of clean tools for that, but it's a pretty simple matter to initialize a new run based on a plot file of a previous run, provided you have a strategy for initializing the state outside of the region of the earlier case. Essentially, you could follow the MFIter loop that calls |
Beta Was this translation helpful? Give feedback.
-
Another quick (and naïve) question on External BCs. Let's say I have periodic BCs on x and y planes, and external boundary functions on z.lo and z.hi, which are different functions. Based on my limited understanding, the FillPatch routines let me use a pointer function to apply external BCs that is being called for the boxes at the boundary. What is the best approach to handle more than one external boundary function without combining them together? Thank you again. |
Beta Was this translation helpful? Give feedback.
I don't know of clean tools for that, but it's a pretty simple matter to initialize a new run based on a plot file of a previous run, provided you have a strategy for initializing the state outside of the region of the earlier case. Essentially, you could follow the MFIter loop that calls
init_data
with some code that reads the plot file (of the solution on the smaller domain) and overwrites-on-intersect. It's a hack, but not terribly ugly.