[Apbs-users] question about grid resolution?

[<vin...@dr...>]

Hi ---

I had a question about electrostatic energies as computed by APBS.

In APBS, the electrostatic energy is defined as the integral over the 
simulation volume of the following density:

F = (q_f/e)u - (diel * kT / 8 * pi * e^2) (\grad u)^2
      - \sum_i c_i [exp(-z_i u) - 1]

where the first term is the interaction of the fixed charge with the 
dimensionless potential u, the second term is the integral of the 
magnitude of the electric field (grad u), and the last term is the sum of 
the local ion excess (or deficit) for each species i.

My question is:

For a fixed amount of memory, is it better to have a finer grid or a 
larger simulation volume?

It seems that if you go with a large volume (but coarser grid) you'll have 
more accuracy for the last term (the ion excess/deficit) .. but perhaps at 
the expense of the first two terms.

Which part of the free energy density is more important?

Thanks,

Vince