My Community

Hello, I am running VASP 5.3 and have been trying to run metadynamics, but cannot seem to produce HILLSPOT or REPORT files just to see if I am indeed producing a bias. I tried both andersen and now Nose-Hoover thermostat. Just for test case I have reduce the HILLS_BIN to 2, and NSW to 5, which if I understand correctly means Gaussians will be deposited for bias potential every two MD steps and total 5 MD steps will take place so should produce 2 depositions (I am not worried about suitability of these parameters, just want to see some output). The MD simulation seems to run fine with OUTCAR showing all 5 steps and no errors. However, I do not find any HILLSPOT file in the same working directory and cannot be sure if metadynamics is indeed running. I had tried one constraint and one bias for the ICONST file, now I am trying two biases. I would appreciate any help to see if I am making any obvious mistake. I am not using a PENALTYPLOT file for this initial run.

MY ICONST file is the following:
R 49 110 5
A 44 43 49 5

My INCAR file is the following:
System ='MD'
LCHARG = .FALSE. ; LWAVE = .FALSE.
PREC = Normal
LREAL= Auto
NCORE= 8
ALGO = very fast

IBRION = 0 ; NSW = 5 ; POTIM = 1.0
ISIF = 2
TEBEG = 500; TEEND = 500
SMASS = -1
MDALGO=21

#BMIX = 0.0001
MAXMIX = 40
ISYM = 0
NWRITE = 0
ISMEAR = 1; SIGMA = 0.172;
NELMIN=4
HILLS_H=0.01
HILLS_W=0.01
HILLS_BIN=2

LBLUEOUT = .TRUE.

Hello, I tried adding this to the INCAR file and still did not get the output files, I also tried it on a different system with VASP 5.2 and same result of no HILLSPOT or REPORT

This is for anyone who is having the same problem, apparently you have to compile vasp with the -Dtbdyn tag for metadynamics to work so check that. For someone using an already installed copy of vasp, this might not be so obvious. That's what fixed it for us.