Extracting arbitrary step from Dimer run?

[vshapovalov]

Hi!
My Dimer run has been progressing nicely, and then it diverged because of SCF convergence problems. The NEWMODECAR still gives a reasonable reaction mode, but the geometry is significantly out of wack. If I wanted to get cartesian coordinates from OUTCAR corresponding to CENTCAR for, say step 17, where exactly are they printed? Would it be possible to save both CENTCAR and MODECAR for each step in a file similar to XDATCAR?

Thank you-
Vladimir

[graeme]

You can extract the central point from the OUTCAR, but not the mode at that step. To get the CENTCAR data, look in the OUTCAR for the line:

Dimer: Central Point

nearest to the iteration of interest. A few lines before that will have the geometry of the center of the dimer.