Fwd: my vc-relax is not converging any help please?
KK M
From: KK M <kkmwo...@gmail.com>
Date: Tue, Dec 16, 2025 at 4:46 PM
Subject: my vc-relax is not converging any help please?
To: <us...@lists.quantum-espresso.org>
I am performing a vc-relax calculation following an initial scf run. Below are the relevant vc-relax input parameters (excluding the CELL_PARAMETERS).
While the total force is being monitored and shows an overall decreasing trend, it continues to fluctuate and does not converge. I have tested different optimisation algorithms (bfgs, cg) as well as various smearing schemes (cold, gaussian), but none of these approaches have led to convergence. As a result, a significant amount of wall time has been consumed without success.
Could you please advise on possible reasons for the lack of convergence in this vc-relax calculation?
For context, the system is a perovskite slab with a water molecule adsorbed on the top surface. The c-direction contains substantial vacuum, as this is effectively a surface calculation. A similar setup with oxygen adsorption converged successfully using comparable parameters, but the present system does not.
&CONTROL
calculation = 'vc-relax'
etot_conv_thr = 1.9000000000d-04
forc_conv_thr = 1.0000000000d-04
outdir = './out/'
prefix = 'In_Sn'
pseudo_dir = '../../pseudos/'
tprnfor = .true.
tstress = .true.
nstep = 220
verbosity = 'high'
max_seconds = 14400
restart_mode = 'restart'
/
&SYSTEM
degauss = 2.0000000000d-02
ecutrho = 600
ecutwfc = 60
ibrav = 0
nat = 25
nosym = .false.
ntyp = 6
A = 7.5000
occupations = 'smearing'
smearing = 'gauss'
/
&ELECTRONS
conv_thr = 1.000000000d-09
electron_maxstep = 180
mixing_beta = 4.0000000000d-01
/
&IONS
ion_dynamics = 'bfgs'
/
&CELL
cell_dynamics = 'bfgs'
press = 0.00
cell_factor = 2.0
press_conv_thr = 0.5
/
Cl 35.453 cl_pbe_v1.4.uspp.F.UPF
Cs 132.9054519 Cs_pbe_v1.uspp.F.UPF
In 114.818 In.pbe-dn-rrkjus_psl.0.2.2.UPF
Sn 118.71 Sn_pbe_v1.uspp.F.UPF
O 15.99900 O.pbe-n-kjpaw_psl.0.1.UPF
H 1.00750 H.pbe-rrkjus_psl.1.0.0.UPF