Return to top-level of LAMMPS documentation
This file outlines the force-field formulas used in LAMMPS. Read this file in conjunction with the data_format and units file.
The sections of this page are as follows:
Whatever Coulomb style is specified in the input command file, the short-range Coulombic interactions are computed by this formula, modified by an appropriate smoother for the smooth, Ewald, and PPPM styles.
E = C q1 q2 / (epsilon * r)
r = distance (computed by LAMMPS)
C = hardwired constant to convert to energy units
q1,q2 = charge of each atom in electron units (proton = +1),
specified in "Atoms" entry in data file
epsilon = dielectric constant (vacuum = 1.0),
set by user in input command file
The style of nonbond potential is specified in the input command file.
E = 4 epsilon [ (sigma/r)^12 - (sigma/r)^6 ] standard Lennard Jones potential r = distance (computed by LAMMPS) coeff1 = epsilon (energy) coeff2 = sigma (distance) 2 coeffs are listed in data file or set in input script 1 cutoff is set in input script
E = 4 epsilon [ (sigma/r)^12 - (sigma/r)^6 ] for r < r_inner
= spline fit for r_inner < r < cutoff
= 0 for r > cutoff
switching function (spline fit) is applied to standard LJ
within a switching region (from r_inner to cutoff) so that
energy and force go smoothly to zero
spline coefficients are computed by LAMMPS
so that at inner cutoff (r_inner) the potential, force,
and 1st-derivative of force are all continuous,
and at outer cutoff (cutoff) the potential and force
both go to zero
r = distance (computed by LAMMPS)
coeff1 = epsilon (energy)
coeff2 = sigma (distance)
2 coeffs are listed in data file or set in input script
2 cutoffs (r_inner and cutoff) are set in input script
E = 4 epsilon [ (sigma/(r - delta))^12 - (sigma/(r - delta))^6 ] same as lj/cutoff except that r is shifted by delta r = distance (computed by LAMMPS) coeff1 = epsilon (energy) coeff2 = sigma (distance) coeff3 = delta (distance) 3 coeffs are listed in data file or set in input script 1 cutoff is set in input script
E = A * [ 1 + cos( pi * r / cutoff ) ] useful for pushing apart overlapping atoms by ramping A over time r = distance (computed by LAMMPS) coeff1 = prefactor A at start of run (energy) coeff2 = prefactor A at end of run (energy) 2 coeffs are listed in data file or set in input script 1 cutoff is set in input script
E = epsilon [ 2 (sigma/r)^9 - 3 (sigma/r)^6 ] used with class2 bonded force field r = distance (computed by LAMMPS) coeff1 = epsilon (energy) coeff2 = sigma (distance) 2 coeffs are listed in data file or set in input script 1 cutoff is set in input script
The coefficients for each nonbond style are input in either the data file by the "read data" command or in the input script using the "nonbond coeff" command. In the former case, only one set of coefficients is input for each atom type. The cross-type coeffs are computed using one of three possible mixing rules:
geometric: epsilon_ij = sqrt(epsilon_i * epsilon_j)
sigma_ij = sqrt(sigma_i * sigma_j)
arithmetic: epsilon_ij = sqrt(epsilon_i * epsilon_j)
sigma_ij = (sigma_i + sigma_j) / 2
sixthpower: epsilon_ij =
(2 * sqrt(epsilon_i*epsilon_j) * sigma_i^3 * sigma_j^3) /
(sigma_i^6 + sigma_j^6)
sigma_ij= ((sigma_i**6 + sigma_j**6) / 2) ^ (1/6)
The default mixing rule for nonbond styles lj/cutoff, lj/switch, lj/shift, and soft is "geometric". The default for nonbond style class2/cutoff is "sixthpower".
The default can be overridden using the "mixing style" command. The one exception to this is for the nonbond style soft, for which only an epsilon prefactor is input. This is always mixed geometrically.
Also, for nonbond style lj/shift, the delta coefficient is always mixed using the rule
The style of bond potential is specified in the input command file.
E = K (r - r0)^2 standard harmonic spring r = distance (computed by LAMMPS) coeff1 = K (energy/distance^2) (the usual 1/2 is included in the K) coeff2 = r0 (distance) 2 coeffs are listed in data file or set in input script
E = -0.5 K R0^2 * ln[1 - (r/R0)^2] +
4 epsilon [(sigma/r)^12 - (sigma/r)^6] + epsilon
finite extensible nonlinear elastic (FENE) potential for
polymer bead-spring models
see Kremer, Grest, J Chem Phys, 92, p 5057 (1990)
r = distance (computed by LAMMPS)
coeff1 = K (energy/distance^2)
coeff2 = R0 (distance)
coeff3 = epsilon (energy)
coeff4 = sigma (distance)
1st term is attraction, 2nd term is repulsion (shifted LJ)
1st term extends to R0
2nd term only extends to the minimum of the LJ potential,
a cutoff distance computed by LAMMPS (2^(1/6) * sigma)
4 coeffs are listed in data file or set in input script
E = -0.5 K R0^2 * ln[1 - ((r - delta)/R0)^2] +
4 epsilon [(sigma/(r - delta))^12 - (sigma/(r - delta))^6] + epsilon
same as FENE/standard expect that r is shifted by delta
r = distance (computed by LAMMPS)
coeff1 = K (energy/distance^2)
coeff2 = R0 (distance)
coeff3 = epsilon (energy)
coeff4 = sigma (distance)
coeff5 = delta (distance)
1st term is attraction, 2nd term is repulsion (shifted LJ)
1st term extends to R0
2nd term only extends to the minimum of the LJ potential,
a cutoff distance computed by LAMMPS (2^(1/6) * sigma + delta)
5 coeffs are listed in data file or set in input script
E = epsilon (r - r0)^2 / [ lamda^2 - (r - r0)^2 ]
non-harmonic spring of equilibrium length r0
with finite extension of lamda
see Rector, Van Swol, Henderson, Molecular Physics, 82, p 1009 (1994)
r = distance (computed by LAMMPS)
coeff1 = epsilon (energy)
coeff2 = r0 (distance)
coeff3 = lamda (distance)
3 coeffs are listed in data file or set in input script
E = K2 (r - r0)^2 + K3 (r - r0)^3 + K4 (r - r0)^4 r = distance (computed by LAMMPS) coeff1 = r0 (distance) coeff2 = K2 (energy/distance^2) coeff3 = K3 (energy/distance^3) coeff4 = K4 (energy/distance^4) 4 coeffs are listed in data file - cannot be set in input script
The style of angle potential is specified in the input command file.
E = K (theta - theta0)^2 theta = radians (computed by LAMMPS) coeff1 = K (energy/radian^2) (the usual 1/2 is included in the K) coeff2 = theta0 (degrees) (converted to radians within LAMMPS) 2 coeffs are listed in data file
E = K2 (theta - theta0)^2 + K3 (theta - theta0)^3 +
K4 (theta - theta0)^4
theta = radians (computed by LAMMPS)
coeff1 = theta0 (degrees) (converted to radians within LAMMPS)
coeff2 = K2 (energy/radian^2)
coeff3 = K3 (energy/radian^3)
coeff4 = K4 (energy/radian^4)
4 coeffs are listed in data file
The style of dihedral potential is specified in the input command file.
E = K [1 + d * cos (n * phi) ]
phi = radians (computed by LAMMPS)
coeff1 = K (energy)
coeff2 = d (always +1 or -1)
coeff3 = n (1,2,3,4,6)
Cautions when comparing to other force fields:
some force fields reverse the sign convention on d so that
E = K [1 - d * cos(n*phi)]
some force fields divide/multiply K by the number of multiple
torsions that contain the j-k bond in an i-j-k-l torsion
some force fields let n be positive or negative which
corresponds to d = 1,-1
in the LAMMPS force field, the trans position = 180 degrees, while
in some force fields trans = 0 degrees
3 coeffs are listed in data file
E = SUM(n=1,3) { K_n [ 1 - cos( n*Phi - Phi0_n ) ] }
phi = radians (computed by LAMMPS)
coeff1 = K_1 (energy)
coeff2 = Phi0_1 (degrees) (converted to radians within LAMMPS)
coeff3 = K_2 (energy)
coeff4 = Phi0_2 (degrees) (converted to radians within LAMMPS)
coeff5 = K_3 (energy)
coeff6 = Phi0_3 (degrees) (converted to radians within LAMMPS)
6 coeffs are listed in data file
The style of improper potential is specified in the input command file.
E = K (chi - chi0)^2
chi = radians (computed by LAMMPS)
coeff1 = K (energy/radian^2) (the usual 1/2 is included in the K)
coeff2 = chi0 (degrees) (converted to radians within LAMMPS)
in data file, listing of 4 atoms requires atom-1 as central atom
some force fields (AMBER,Discover) have atom-2 as central atom - it is really
an out-of-plane torsion, may need to treat as dihedral in LAMMPS
2 coeffs are listed in data file
same formula, coeffs, and meaning as "harmonic" except that LAMMPS
averages all 3 angle-contributions to chi
in class II this is called a Wilson out-of-plane interaction
2 coeffs are listed in data file
If class II force fields are selected in the input command file, additional cross terms are computed as part of the force field.
E = K (r - r0) * (r' - r0') r,r' = distance (computed by LAMMPS) coeff1 = K (energy/distance^2) coeff2 = r0 (distance) coeff3 = r0' (distance) 3 coeffs are input in data file
E = K_n (r - r0_n) * (theta - theta0) r = distance (computed by LAMMPS) theta = radians (computed by LAMMPS) coeff1 = K_1 (energy/distance-radians) coeff2 = K_2 (energy/distance-radians) coeff3 = r0_1 (distance) coeff4 = r0_2 (distance) Note: theta0 is known from angle coeffs so don't need it specified here 4 coeffs are listed in data file
E = (r - r0) * [ F1*cos(phi) + F2*cos(2*phi) + F3*cos(3*phi) ] r = distance (computed by LAMMPS) phi = radians (computed by LAMMPS) coeff1 = F1 (energy/distance) coeff2 = F2 (energy/distance) coeff3 = F3 (energy/distance) coeff4 = r0 (distance) 4 coeffs are listed in data file
E = (r - r0_n) * [ F1_n*cos(phi) + F2_n*cos(2*phi) + F3_n*cos(3*phi) ] r = distance (computed by LAMMPS) phi = radians (computed by LAMMPS) coeff1 = F1_1 (energy/distance) coeff2 = F2_1 (energy/distance) coeff3 = F3_1 (energy/distance) coeff4 = F1_2 (energy/distance) coeff5 = F2_3 (energy/distance) coeff6 = F3_3 (energy/distance) coeff7 = r0_1 (distance) coeff8 = r0_2 (distance) 8 coeffs are listed in data file
E = (theta - theta0) * [ F1_n*cos(phi) + F2_n*cos(2*phi) + F3_n*cos(3*phi) ] theta = radians (computed by LAMMPS) phi = radians (computed by LAMMPS) coeff1 = F1_1 (energy/radians) coeff2 = F2_1 (energy/radians) coeff3 = F3_1 (energy/radians) coeff4 = F1_2 (energy/radians) coeff5 = F2_3 (energy/radians) coeff6 = F3_3 (energy/radians) coeff7 = theta0_1 (degrees) (converted to radians within LAMMPS) coeff8 = theta0_2 (degrees) (converted to radians within LAMMPS) 8 coeffs are listed in data file
E = K (theta - theta0) * (theta' - theta0') * (phi - phi0) theta,theta' = radians (computed by LAMMPS) phi = radians (computed by LAMMPS) coeff1 = K (energy/radians^3) coeff2 = theta0 (degrees) (converted to radians within LAMMPS) coeff3 = theta0' (degrees) (converted to radians within LAMMPS) Note: phi0 is known from dihedral coeffs so don't need it specified here 3 coeffs are listed in data file
(undocumented)
E = K_n (theta - theta0_n) * (theta' - theta0_n') theta,theta' = radians (computed by LAMMPS) coeff1 = K_1 (energy/radians^2) coeff2 = K_2 (energy/radians^2) coeff3 = K_3 (energy/radians^2) coeff4 = theta0_1 (degrees) (converted to radians within LAMMPS) coeff5 = theta0_2 (degrees) (converted to radians within LAMMPS) coeff6 = theta0_3 (degrees) (converted to radians within LAMMPS) 6 coeffs are listed in data file