RadialSolver: Dynamic Liquid Layer Stability Issues

[jrenaud90]

Describe the bug
This is a thread to track improvements / issues related to integration stability when using dynamic liquid layers in RadialSolver.

• Causes
  • Stability problems appear when a dynamic liquid layer is present in a world with a moderately large forcing period.
  • The value of the forcing period depends on the density of the liquid. A low density liquid tends to be more stable at higher forcing periods than a high density layer.

Note that "forcing period" refers to the tidal mode that is being calculated. A planet with a low orbital period may still require larger forcing period modes depending on its eccentricity, obliquity, and rotation rate.

[jrenaud90]

The new RadialSolver has a warning when provided with a high forcing period and a liquid layer. Currently this warning does not check for the density of the layer.

It would be nice to come up with a stability term (function of frequency, density and ???) that can be used in these warning checks.