[ESPResSo-devel] p3m kspace force calc, p3m stress tensor

[Tobias Richter]

Hello,

I am currently trying to repair the contributions to the stress tensor,
arising from the longrange electrostatics, so the k-space part.
For this I essentially need the forces on the mesh, or more precisely the
electric field, since I evaluate the Maxwellian stress tensor.

The problem I'm facing now is, that some components in the electric field
still have non negligible imaginary components, which should theoretically
be zero, since the acting force has to be real valued. I find those
nonzero imaginary parts in the p3m_calc_kspace_forces function, before the
forces get assigned to the grid, so when the mesh should contain the
electric field, which (as I believe) should be strictly real valued.

So is this just my misunderstanding or is there something wrong in the
calculation?

Kind regards,
Tobias