MD

This section defines the whole set of parameters needed perform an MD run. [Edit on GitHub]

Keywords

Keyword descriptions

ANGVEL_TOL: real = [au_t^-1*bohr]

Usage: angvel_tol 0.1

The maximum accepted angular velocity. This option is ignored when the system is periodic. Removes the components of the velocities that project on the external rotational degrees of freedom. [Edit on GitHub]

ANGVEL_ZERO: logical = F

Lone keyword: T

Usage: angvel_zero LOGICAL

Set the initial angular velocity to zero. This option is ignored when the system is periodic or when initial velocities are defined. Technically, the part of the random initial velocities that projects on the external rotational degrees of freedom is subtracted. [Edit on GitHub]

ANNEALING: real = 1.00000000E+000

Usage: annealing

Specifies the rescaling factor for annealing velocities. Automatically enables the annealing procedure. This scheme works only for ensembles that do not have thermostats on particles. [Edit on GitHub]

ANNEALING_CELL: real = 1.00000000E+000

Usage: ANNEALING_CELL

Specifies the rescaling factor for annealing velocities of the CELL Automatically enables the annealing procedure for the CELL. This scheme works only for ensambles that do not have thermostat on CELLS velocities. [Edit on GitHub]

COMVEL_TOL: real = [au_t^-1*bohr]

Usage: comvel_tol 0.1

The maximum accepted velocity of the center of mass. With Shell-Model, comvel may drift if MD%THERMOSTAT%REGION /= GLOBAL [Edit on GitHub]

DISPLACEMENT_TOL: real = 5.29177209E+001 [angstrom]

Usage: DISPLACEMENT_TOL

This keyword sets a maximum atomic displacement in each Cartesian direction. The maximum velocity is evaluated and if it is too large to remain within the assigned limit, the time step is rescaled accordingly, and the first half step of the velocity verlet is repeated. [Edit on GitHub]

ECONS_START_VAL: real = 0.00000000E+000 [hartree]

Usage: econs_start_val

The starting value of the conserved quantity [Edit on GitHub]

ENSEMBLE: enum = NVE

Usage: ensemble nve

Valid values:

  • NVE constant energy (microcanonical)

  • NVT constant temperature and volume (canonical)

  • NPT_I constant temperature and pressure using an isotropic cell

  • NPT_F constant temperature and pressure using a flexible cell

  • MSST simulate steady shock (uniaxial)

  • MSST_DAMPED simulate steady shock (uniaxial) with extra viscosity

  • HYDROSTATICSHOCK simulate steady shock with hydrostatic pressure

  • ISOKIN constant kinetic energy

  • REFTRAJ reading frames from a file called reftraj.xyz (e.g. for property calculation)

  • LANGEVIN langevin dynamics (constant temperature)

  • NPE_F constant pressure ensemble (no thermostat)

  • NPE_I constant pressure ensemble using an isotropic cell (no thermostat)

  • NVT_ADIABATIC adiabatic dynamics in constant temperature and volume ensemble (CAFES)

  • NPT_IA NPT_I ensemble with frozen atoms in absolute coordinate

References: EVANS1983, VandeVondele2002, Minary2003

Mentions:QM/MM with Built-in Force Field, ⭐Langevin Dynamics

The ensemble/integrator that you want to use for MD propagation [Edit on GitHub]

INITIALIZATION_METHOD: enum = DEFAULT

Usage: INITIALIZATION_METHOD DEFAULT

Valid values:

  • DEFAULT Assign random velocities and then scale according to TEMPERATURE

  • VIBRATIONAL Initialise positions and velocities to give canonical ensemble with TEMPERATURE, using the method described in PRL 96, 115504 (2006)

This keyword selects which method to use to initialize MD. If velecities are not set explicitly, DEFAULT optioin will assign random velocities and then scale according to TEMPERATURE; VIBRATIONAL option will then use previously calculated vibrational modes to initialise both the atomic positions and velocities so that the starting point for MD is as close to canonical ensemble as possible, without the need for lengthy equilibration steps. See PRL 96, 115504 (2006). The user input atomic positions in this case are expected to be already geometry optimised. Further options for VIBRATIONAL mode is can be set in INITIAL_VIBRATION subsection. If unspecified, then the DEFAULT mode will be used. [Edit on GitHub]

MAX_STEPS: integer = 1000000000

Usage: max_steps 100

The number of MD steps to perform, counting from step 1 [Edit on GitHub]

SCALE_TEMP_KIND: logical = F

Lone keyword: T

Usage: scale_temp_kind LOGICAL

When necessary rescale the temperature per each kind separately [Edit on GitHub]

STEPS: integer = 3

Usage: steps 100

Mentions:X-Ray Resonant Excitation

The number of MD steps to perform, counting from step_start_val. [Edit on GitHub]

STEP_START_VAL: integer = 0

Usage: step_start_val

The starting step value for the MD [Edit on GitHub]

TEMPERATURE: real = 3.00000000E+002 [K]

Usage: temperature 325.0

Mentions:Langevin Dynamics

The temperature in K used to initialize the velocities with init and pos restart, and in the NPT/NVT simulations [Edit on GitHub]

TEMPERATURE_ANNEALING: real = 1.00000000E+000

Usage: TEMPERATURE_ANNEALING

Specifies the rescaling factor for the external temperature. This scheme works only for the Langevin ensemble. [Edit on GitHub]

TEMP_KIND: logical = F

Lone keyword: T

Usage: temp_kind LOGICAL

Compute the temperature per each kind separately [Edit on GitHub]

TEMP_TOL: real = 0.00000000E+000 [K]

Aliases: TEMP_TO ,TEMPERATURE_TOLERANCE

Usage: temp_tol 0.0

Mentions:Molecular Dynamics

The maximum accepted deviation of the (global) temperature from the desired target temperature before a rescaling of the velocites is performed. If it is 0 no rescaling is performed. NOTE: This keyword is obsolescent; Using a CSVR thermostat with a short timeconstant is recommended as a better alternative. [Edit on GitHub]

TIMESTEP: real = 5.00000000E-001 [fs]

Usage: timestep 1.0

Mentions:X-Ray Absorption from δ-Kick, ⭐X-Ray Resonant Excitation, ⭐QM/MM with Built-in Force Field

The length of an integration step (in case RESPA the large TIMESTEP) [Edit on GitHub]

TIME_START_VAL: real = 0.00000000E+000 [fs]

Usage: time_start_val

The starting timer value for the MD [Edit on GitHub]