LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] compute ti command
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] compute ti command

From: Agilio Padua <agilio.padua@...24...>
Date: Sat, 5 May 2018 20:03:37 +0200


It may be because the lambda parameter in the soft-core-softed potentials has a more complicated functional form and so your analytical derivative df(lambda)/dlambda is not 1.0
Mixing lj/cut/soft with compute ti is not simple…
compute fep was made to work with the core-softer potentials, so it should work fine.

Maybe you can test compute ti against the other methods on a free energy route that does not involve starting from 0 interactions and in this case you don’t need the soft potentials.

Hope this helps,


I have been trying to test a few free energy methods to calculate the solvation free energy of a LJ solute (B) in the same LJ solvent (A),  a very simple model system.
These methods include free energy perturbation (FEP), Widom insertion, and thermodynamic integration (TI) methods.

The results I've got are as following:
FEP:     -3.61 (0.05)
Widom:             -3.78 (0.2)
TI:  -4.22 (0.02)

I can't see why TI method doesn't give the right answer. Here is my input file:

units lj
dimension 3
boundary p p p
atom_style atomic

pair_style hybrid    lj/cut 2.5  lj/cut/soft 1.0 0.5 2.5 
mass   1      1.0000 # A
mass   2      1.0000 # B
pair_coeff  1     1    lj/cut            1.0 1.0 # A    A
pair_coeff  1     2     lj/cut/soft 1.0 1.0  ${Lambda} # A    B 
pair_coeff  2     2    lj/cut            1.0 1.0 # B    B 

restart         10000

thermo 10000
timestep        0.01
run_style       verlet

variable lambda equal 1
variable           dlambda equal 1

compute TI all  ti lj/cut/soft   1*2   v_lambda v_dlambda

fix   NPT all npt temp 0.8 0.8 1.0  iso 0.1  0.1 10.0 
run 300000

reset_timestep 0
fix thermodynamic all ave/time 1 100 100   c_TI    file   Thermodynamic.dat

run 1000000

Please note that I am doing separate independent simulations in a parallel way by having ${Lambda} from 0 to 1.0 with 0.025 intervals (41 simulations). 
I'm using lammps 16Mar18. 
Any thoughts will be appreciated.