diff options
| author | Eugeniy Mikhailov <evgmik@gmail.com> | 2012-06-13 17:19:50 -0400 |
|---|---|---|
| committer | Eugeniy Mikhailov <evgmik@gmail.com> | 2012-06-13 17:19:50 -0400 |
| commit | f45eba3c8ea3f613fba3fc7dc67bd13a45f74006 (patch) | |
| tree | 1429d41b5842973d21c417b8b86bc7e5ab21eaed /xmds2 | |
| parent | 7f6bec7c236e2d1e17000d0d44172404161e784f (diff) | |
| download | Nresonances-f45eba3c8ea3f613fba3fc7dc67bd13a45f74006.tar.gz Nresonances-f45eba3c8ea3f613fba3fc7dc67bd13a45f74006.zip | |
changed time lattice to preserve convergence at large detuning
Diffstat (limited to 'xmds2')
| -rw-r--r-- | xmds2/Nlevels_with_doppler_with_z_4wm/Nlevels_with_doppler_with_z_4wm.xmds | 12 |
1 files changed, 11 insertions, 1 deletions
diff --git a/xmds2/Nlevels_with_doppler_with_z_4wm/Nlevels_with_doppler_with_z_4wm.xmds b/xmds2/Nlevels_with_doppler_with_z_4wm/Nlevels_with_doppler_with_z_4wm.xmds index 054d343..36d47a9 100644 --- a/xmds2/Nlevels_with_doppler_with_z_4wm/Nlevels_with_doppler_with_z_4wm.xmds +++ b/xmds2/Nlevels_with_doppler_with_z_4wm/Nlevels_with_doppler_with_z_4wm.xmds @@ -136,7 +136,17 @@ <geometry> <propagation_dimension> z </propagation_dimension> <transverse_dimensions> - <dimension name="t" lattice="1000" domain="(-1e-6, 1e-6)" /> + <!-- IMPORTANT: looks like having a lot of points in time helps with convergence. + I suspect that time spacing should be small enough to catch + all pulse harmonics and more importantly 1/dt should be larger than + the largest detuning (including Doppler shifts). + Unfortunately calculation time is proportional to lattice size + so we cannot just blindly increase it. + Some rules of thumb: + * lattice="1000" domain="(-1e-6, 1e-6)" + was good enough detunings up to 1.5GHz=9.8rad/s notice that 1/dt=500MHz + --> + <dimension name="t" lattice="5000" domain="(-1e-6, 1e-6)" /> <dimension name="v" lattice="100" domain="(V_maxwell_min, V_maxwell_max)" /> </transverse_dimensions> </geometry> |