[DFTB-Plus-User] Problem with SCC correction when calling DFTB+ externally via ASE

[Robin Shannon]

Hi


I have been calling DFTB+ as an external forcefield for use in my own trajectory code utilising the ASE python package. When running DFTB+ without an SCC correction all appears normal however this changes when I apply an SCC correction. For example I have a number of small molecule input files for which  geometry optimisations  converge readily within DFTB+ but which atomise when I use the DFTB+ forces with the ASE optimisers. This is not due to a problem with the ASE optimisers since without the SCC correction both DFTB+ and ASE give identical results.


I assume this something todo with they way DFTB+ uses converged SCC solutions from a previous step but if anyone has advice on using DFTB+ as a external forcefield I would be grateful.


Many thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20171122/d80bdd87/attachment.html>