Lagrangian metadynamics parameters & tuning

98 views
Skip to first unread message

alh chin

unread,
Nov 1, 2022, 8:27:25 AM11/1/22
to cp2k
Dear all, 

I want to use Lagrangian MTD to study a reaction (since the barrier appears to be high), I checked the CP2K manuals and the example at exercises:2015_cecam_tutorial:mtd1 [CP2K Open Source Molecular Dynamics ]
I found that I need to specify 4 parameters for Lagrangian MTD, 
(1)TEMPERATURE
(2)TEMP_TOL  
and for each CV, 
(3) LAMBDA
(4) MASS

my questions are 
(1) I understand why TEMPERATURE need to be specified, but why TEMP_TOL (temperature tolerance) has to be specified?
(2) is LAMBDA the force constant for the fictitious particle? what's the unit?
(3) what would be good values for TEMP_TOL and (most importantly) LAMBDA, and, how to tune LAMBDA? 

Thanks,

Alh.



Marcella Iannuzzi

unread,
Nov 1, 2022, 9:41:03 AM11/1/22
to cp2k
Dear Alh.

If not specified, the temperature is set to the system's temperature
The tolerance for the temperature can but must not be specified. 
The units of the force constant are atomic units.
The choice of all these parameters is system dependent.  
Preparatory runs to explore the system's behaviour are strongly recommended to help in choosing the proper settings.

Regards
Marcella

alh chin

unread,
Nov 2, 2022, 4:21:41 AM11/2/22
to cp...@googlegroups.com
Dear Marcella, 

Thank you very much for your reply.  I found a few papers that discuss the tuning of the force constants, e.g., Ensing et al., (J. Phys. Chem. B 2005, 109, 6676-6687). From what I understand, the force constant has to be tuned such that the path of the fictitious particle is not that far away from the real CV path. So I need to prepare some data of the fictitious particle (with test runs). I would expect these data to be stored in a file called XXX_COLVAR.metadynLog.

In the example you explained that columns in the metadynLog file are CVs, gradients of the bias potential wrt CVs, gradients of the wall wrt CVs, values of the bias potentials, values of the wall potentials.

I would expect if a Lagrangian MTD is run, the coordinates of fictitious particle would be monitored in the metadynLog.

QUESTION:  what are the columns in the metadynLog file for a Lagrangian MTD

thanks, 

Alh,

------------------

(the example you provided is below, I would guess the final one is temp)

         0.0     2.71770373     1.42316747     0.11550959     2.71061596     1.42089234     0.11424663     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00021431     0.00006879     0.00003819     0.00000000     0.00000000     0.00000000   300.00000000
         0.8     2.72482415     1.42536380     0.11672755     2.71045717     1.41982383     0.11363373    -0.01811640    -0.01003091    -0.00562756     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00021529     0.00006641     0.00003683     0.00008769     0.00000000     0.00000000   300.00000000
         1.6     2.73206759     1.42720809     0.11779474     2.71219970     1.41629628     0.11236277    -0.03156113    -0.02720256    -0.01309435     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00021901     0.00005576     0.00003227     0.00035113     0.00000000     0.00000000   300.00000000


with the input file:


&FREE_ENERGY
    &METADYN
      DO_HILLS 
      NT_HILLS 100
      WW 2.0e-3
      LAGRANGE
      TEMPERATURE 300.
      TEMP_TOL  10.
      &METAVAR
        LAMBDA  2.5
        MASS   30.
        SCALE 0.1
        COLVAR 1
      &END METAVAR
      &METAVAR
        LAMBDA 3.0
        MASS 30.0
        SCALE 0.25
        COLVAR 2
      &END METAVAR
      &METAVAR
        LAMBDA 3.0
        MASS 30.0
        SCALE 0.25
        COLVAR 3
        &WALL
            POSITION 0.0
            TYPE QUARTIC
            &QUARTIC
               DIRECTION WALL_MINUS
               K  100.0
            &END
        &END
      &END METAVAR

      &PRINT
        &COLVAR
           COMMON_ITERATION_LEVELS 3
           &EACH
             MD 1
           &END
        &END
        &HILLS
           COMMON_ITERATION_LEVELS 3
           &EACH
             MD 1
           &END
        &END
      &END
    &END METADYN
  &END

  &PRINT
    &TRAJECTORY
      &EACH
        MD 50
      &END
    &END
    &VELOCITIES OFF
    &END
    &RESTART
      &EACH
         MD 100
      &END
      ADD_LAST NUMERIC
    &END
    &RESTART_HISTORY 
      &EACH
         MD 2000
      &END
    &END
  &END
&END MOTION

--
You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/tvrAI_GiBBY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/cb19fb3c-cc78-4fe7-b734-2c9f4d66d01cn%40googlegroups.com.

Marcella Iannuzzi

unread,
Nov 2, 2022, 4:49:12 AM11/2/22
to cp2k



For the specific case

time     metavar1        metavar2          metavar3             CV1                   CV2                 CV3                    FH1                    FH2                  FH3                  MTDF1               MTDF2              MTDF3             WF1                     WF2                     WF3                VEL1       VEL2                   VEL3                    HPOT           MTDPOT              WALLPOT        TEMP
0.0     2.71770373     1.42316747     0.11550959     2.71061596     1.42089234     0.11424663     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00000000     0.00021431     0.00006879     0.00003819     0.00000000     0.00000000     0.00000000   300.00000000


where FH is the force contribution on metavar from the harmonic potential, MTDF is the force contribution from the MTD potential, WF is the force contribution from the wall potential

The rest should be selfexplaining.
Regards
Marcella
Reply all
Reply to author
Forward
0 new messages