Friday, January 13, 2012

[exqukuvh] Modeling a gas

Start by modeling a two dimensional gas in an enclosed rectangle.  This could be done by lots of elastic spheres bouncing off each other, but there are probably easier, more continuous ways.  Initialize some high and low pressure areas, then let the system evolve.  Assuming perfect conservation, there should result standing waves.

Instead of an enclosed box, do the simulation on the surface of a sphere -- no boundary conditions (or, periodic boundary conditions).

Transform the simulation onto a reference frame of a rotating sphere.  So far, this is super unrealistic: the gas is frictionlessly unaffected by the motion of the "ground", so the rotating reference frame just makes things look different.

Next, add a third dimension of altitude and gravity.  Maybe next add an overhead heat source and PV=nRT.

Although this could be further elaborated to a global weather model, one can also stop and enjoy this conceptually simple model so far and its flowing patterns of high and low pressure areas.

No comments :