GEOMETRY wrong or EMAX_SPLINE too small!

[Puneet Chandel]

Hello,

I am running geometry optimization of lattice structure of silicate with one unit cell. I have taken the coordinated from crystallographic database. I use the PBC exactly equal to the edge length given by XRD data so that unit cell can replicate itself. But it gives the error that GEOMETRY wrong or EMAX_SPLINE too small!. Although with  increase in PBC the GO starts but it does not converge and the geometry get distorted because the unit cell do not replicate itself and atoms overlap.

can anyone suggest how to minimize unit cell with PBC equal to edge length.

Thanks for your help

[Krack Matthias]

Remove all atoms equivalent (because of the PBC) to an already defined one, e.g. atom 12 which is equivalent to atom 11. The list of too short distance warning can help you to identify the redundant atoms.

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/3d6046a0-d3a1-4d3f-ba3b-e0cfe273b419n%40googlegroups.com.

[Puneet Chandel]

Dear Matthias,

Thank you for your kind reply. You are absolutely right about removing similar atoms. Now the system is getting converged and the atoms were replicating on the exact positions while applying PBC. But when I visualize the optimized structure using VMD, there are no bonds between the atoms. Can you suggest me something to get bonds between atoms?  

Thanks for your help.

[Krack Matthias]

That’s something you have to adjust in VMD, i.e. you have to define the maximum distance between atoms for drawing bonds in VMD.

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/1f319213-2806-457d-b942-7f2db2a33eb3n%40googlegroups.com.