UofM-Summer-School,

June 25-28, 2018

Introduction to Molecular
Dynamics simulations
Ali Kerrache

E-mail: [email protected]

Summer School, June 25-28, 2018

 Who am I?
q High Performance Computing Analyst
Ø Grex: support for UofM users
Ø WestGrid and Compute Canada.
Ø Software and User Support.
Ø National teams: Grex
ü BST: Bio-molecular Simulation Team.
ü RSNT: Research Support National Team.

q Computational Physicist
Ø Monte Carlo and Molecular Dynamics codes.
Ø Study of the properties of materials using MD simulation.
v Metals, Glasses: Silica, Amorphous silicon, Nuclear Glasses.
v Mass transport, solid-liquid interfaces, kinetic coefficients, melting,
crystallization, mechanical deformations, static and dynamical properties,
He diffusion in glasses, …

Summer School, June 25-28, 2018

 Outline
q Classical Molecular Dynamics simulations:
Ø Introduction
Ø Classical MD: basics
Ø Algorithms and force fields used in MD
Ø Some results:
Ø Crystallization of AlNi
Ø Shear deformation is a-Si
Ø Indentation

q Setting and running MD simulations: LAMMPS:

Ø Introduction to LAMMPS
Ø Building LAMMPS: demonstration
Ø Running LAMMPS: demonstration
Ø Benchmarks and performance tests
Summer School, June 25-28, 2018
 Download the required material
q Use ssh client: PuTTy, MobaXterm, Terminal (Mac or Linux) to
connect to cedar and/or graham:

Ø ssh –Y [email protected]
Ø ssh –Y [email protected]

q Download the files using wget:

wget https://ali-kerrache.000webhostapp.com/uofm/md.tar.gz
wget https://ali-kerrache.000webhostapp.com/uofm/md-slides.pdf

q Unpack the archive and change the directory:

tar -xvf md.tar.gz
cd UofM-Summer-School-MD

Summer School, June 25-28, 2018

 Why do we need simulations?
q Except for simple cases: no analytical
solutions for most of the problems.
Theory
q In most cases, experiments are:
Ø Difficult or impossible to perform.
Ø Too dangerous to …
Ø Expensive and time consuming.
Ø Blind and too many parameters to control.
Simulation

q Simulation is a powerful tool:

Ø can replace some experiments.
Ø provoke experiments.
Experiment
Ø explain and understand experiments.
Ø complete the theory and experiments.

Summer School, June 25-28, 2018

 Atomistic / Molecular simulations
q What are atomistic / molecular Simulation?
Ø a tool to get insights about the properties of materials at atomic
or molecular level.
Ø used to predict and / or verify experiments.
Ø considered as a bridge between theory and experiment.
Ø provide a numerical solution when analytical ones are impossible.
Ø used to resolve the behavior of nature (the physical world
surrounding us) on different time- and length-scales.

q Applications, simulations can be applied in, but not limited to:

ü Physics, Applied Physics, Chemistry, …
ü Materials and Engineering, …
ü and more …

Summer School, June 25-28, 2018

 Length and time scales
v Different methods
Time v Different time scales
v Different length scales
second v Complementarity
Macroscale
microsec.

Mesoscale
nanosec.
Molecular
picosec Dynamics
Quantum
femtosec. Mechanics Length
nanometer micrometer mm meter

Summer School, June 25-28, 2018

 Classical Molecular Dynamics
q Solution of Newton’s equations: ! !
mi ai = Fi
Ø MD is the solution of the classical ! N !
equations of motion for a system of N atoms Fi = å f ij
j ¹i
or molecules in order to obtain the time
evolution of the system.
Ø Uses algorithms to integrate
the equations of motion.
Ø Applied to many-particle systems.
Ø Requires the definition of a force field or
potential to compute the forces. ! !
Ø Potential fitting: first principle calculations f ij = -Ñ iV (rij )
and experiments.
Summer School, June 25-28, 2018
 Structure of MD programs
Initialization ri (t0 ) v i (t0 )

Compute the new forces Fi (ri )

Repeat as necessary

Loop over time/MD steps

ri (t ) ® ri (t + Dt )
Solve the equation of motion vi (t ) ® vi (t + Dt )
v Thermodynamic
Sample / Check point properties, …
v MD trajectory, …

Test and increase time t = t + Dt

t > t max
End of the simulation

Summer School, June 25-28, 2018

 Forces in MD: Newton’s equations
q Potential function:

U (r ) = U bond (...) + U non-bond (...) + U ext (...)

q Evaluate the forces acting on each particle:

v The force on each atom is determined by: Fi = -ÑU (r )

U (r ) : potential function
N : number of atoms in the system
rij : vector distance between atoms i and j
q Newton’s equation of motion:

Summer School, June 25-28, 2018

 Force Fields for MD simulations

Interactions:
Ø Lenard-Jones
Ø Electrostatic
Ø Bonds
Ø Orientation
Ø Rotational

Summer School, June 25-28, 2018

 Derivation of Verlet algorithm
Taylor' s expansions : position acceleration

r (t + Dt ) = r (t ) + r!(t ) Dt + 12 !r!(t ) Dt 2 + 16 !r!!(t ) Dt 3 + O ( Dt 4 ) (I )

r (t - Dt ) = r (t ) - r!(t ) Dt + 12 !r!(t ) Dt 2 - 16 !r!!(t ) Dt 3 + O ( Dt 4 ) (II )
Add (I) and (II) : velocity

r (t + Dt ) + r (t - Dt ) = 2r (t ) + !r!(t ) Dt 2 + O ( Dt 4 )
or :
r (t + Dt ) = 2r (t ) - r (t - Dt ) + f (t ) Dt 2 / m + O ( Dt 4 ) ( II )
Subtract (II) from (I) : {r(t+Dt), v(t+Dt)}
r (t + Dt ) - r (t - Dt ) = 2r!(t ) Dt + O ( Dt 3 )
or :
v(t) = (r (t + Dt ) - r (t - Dt ) ) / 2Dt + O ( Dt 2 ) {r(t), v(t)} ( IV )

Summer School, June 25-28, 2018

 Usage of Verlet algorithm
q From the initial positions and velocities: ri (t ) v i (t )
1
a(r ) = F(r (t ))
m
q Obtain the positions and velocities at: t + Dt
• velocity calculated explicitly 1
• possible to control the temperature r (t + D t ) = r (t ) + v (t ) Dt + a (r ) Dt 2

2
• stable in long simulation
1
• most used algorithm a(t + Dt ) = F (r (t + Dt ))
m
v Leap-Frog algorithm 1
v(t + Dt / 2) = v(t )Dt + a(r )Dt
2
1
v(t + Dt ) = v(t + Dt / 2) + a(t + Dt )Dt
2

Summer School, June 25-28, 2018

 Predictor – Corrector algorithm
q Predictor step: 1
Ø from the initial ri (t ), v i (t ) è a(r ) = F(r (t ))
m
Ø predict ri (t + Dt ) , v i (t + Dt ) using Taylor’s series
a(t ) 2
r (t + Dt ) @ r (t ) + v(t )Dt +
P
Dt
2
v (t + Dt ) @ v(t ) + a(t )Dt
P

iii
a P (t + Dt ) @ a(t ) + r iii (t )Dt r : 3rd order derivatives
F ( r P
(t + Dt ))
q Corrector step: Ø get corrected acceleration: a (r) =
C

m
Ø using error in acceleration: Da(t + Dt ) @ a (t + Dt ) - a (t + Dt )
C P

Dt 2
Ø correct the positions: r(t + Dt ) @ r P (t + Dt ) + C0 Da(t + Dt )
2
Ø correct the velocities: v(t + Dt ) @ v P (t + Dt ) + C1DtDa(t + Dt )
C n: constants depending accuracy

Summer School, June 25-28, 2018

 MD simulation settings
q Starting configuration:
Ø Atomic positions (x,y,z)
Ø density, mass, charge, ….
q Initial velocities: depend on temperature
q boundary conditions (PBC):
Ø PBC: required to simulate bulk properties.
Ø or fixed boundary conditions
q set the appropriate potential:
Ø available and supported potentials
Ø depend on the system to simulate (literature search).
q set the appropriate time step: should be short (order of 1fs).
q set the temperature and pressure control:
Ø define the thermodynamic ensemble (NVT, NPT, NVE, …).
q Fix run time and customize the output: depend on the software.
Summer School, June 25-28, 2018
 Boundary conditions
Ø create images of the simulation box:
duplication in all directions (x, y and z)
Ø if an atom is moving out of boundary,
it comes from the other side.
Ø used also in pair interactions evaluation

Ø PBC:
in x, y directions
Ø Walls:
fixed boundaries
in z direction.

Summer School, June 25-28, 2018

 Neighbor list
Ø Optimization of MD algorithms: For each particle: N-1 pairs.
For N particles: N(N-1) pairs.
Ø Evaluating the forces is time consuming:
v Pair potential calculation: µ O( N 2 )
!
v Atom moves < 0.2 A per time step
rcut
v Cutoff radius: not necessary to include
i
all the possible pairs.
rL
q Solution: Verlet neighbor list
Ø Containing all neighbors of each atom
} skin
within: rL N L vDt
rL - rcut >
Ø Update every N L steps 2
Summer School, June 25-28, 2018
 Thermodynamic ensembles
q Ensembles: Each ensemble is used for a
Ø NVE – micro-canonical ensemble specific simulation:
Ø NVT – canonical ensemble Ø Equilibration, …
Ø NPT – grand-canonical ensemble Ø Production run, …
Ø others … Ø Diffusion (NVE), …

q Temperature control:
Ø Berendsen thermostat (velocity rescaling)
Ø Andersen thermostat
ü Choose the ensemble that best
Ø Nose-Hoover chains fits your system and the properties
you want to simulate
q Pressure control: ü start the simulation.
Ø Berendsen volume rescaling ü Check the thermodynamic
Ø Andersen piston properties as a function of time.

Summer School, June 25-28, 2018

 What is MD used for?
q Goal of MD simulations:
Ø The prime purpose of MD is to sample the phase space of the
statistical mechanics ensemble.
Ø Most physical properties can be related the atomic trajectories and
obtained as average as a function of time.

q Structural properties:
Ø obtained from spatial correlation functions e.g. distribution functions
(RDF, S(Q), Van-Hove, ...).

q Dynamical Properties:
ØTime dependent properties (MSD, diffusion coefficients) obtained via
temporal correlation functions e.g. velocity autocorrelation function,
atomic displacements.

Summer School, June 25-28, 2018

 Thermodynamic properties
1 N
v Kinetic Energy K . E. = å mi vi2
2 i
2
v Temperature T= K .E
3Nk B
N
v Configuration Energy Uc = å å V (rij )
i j >i

1 N -1 N ! !
v Pressure PV = Nk BT - åå rij × f ij
3 i =1 j >i
3 2 2 3 Nk B
v Specific Heat d (Uc ) NVE = Nk B T (1 -
2
)
2 2Cv

Summer School, June 25-28, 2018

 Structural properties
v Radial Distribution Function (simulation)
N
Al25Ni75 n(r ) V
g(r ) = =
4p r r 2 Dr N 2
å å d (r - rij )
i j ¹i

Al80Ni20
¥ sin( kr )
S ( k ) = 1 + 4p r ò (g (r ) - 1) r 2 dr
0 kr
Ø Structure Factor (experiments)
Summer School, June 25-28, 2018
 Dynamical properties
MSD: Mean Square Displacement (Einstein relation)

1
2 Dt = | ri (t ) - ri (0) |2
3

! !
MSD = cAl (rs, Ni (t ) - rs, Ni (0))2
! !
+ cNi (rs, Al (t ) - rs, Al (0))2

Diffusion constants

1 N
(r(t) - r(0))2
D = lim
t ®¥ N
å 6t
i=1

Summer School, June 25-28, 2018

 Recommended textbooks
Ø The Art of Molecular Dynamics Simulation, D.C. Rapaport,
Camb. Univ. Press (2004)
Ø Understanding Molecular Simulation, D. Frenkel and B. Smit,
Academic Press (2002).
Ø Computer Simulation of Liquids, M.P. Allen and D.J. Tildesley,
Oxford (1989).
Ø Theory of Simple Liquids, J.-P. Hansen and I.R. McDonald,
Academic Press (1986).
Ø Classical Mechanics, H. Goldstein, Addison Wesley (1980).
Ø Glassy Materials and Disordered Solids, An Introduction to
their Statistical Mechanics, 2nd edition, Kob, Walter and
Binder, K., 2011

Summer School, June 25-28, 2018

 Available MD codes
q Open source: free access
ü LAMMPS: http://lammps.sandia.gov/index.html
ü DL_POLY: http://www.scd.stfc.ac.uk/SCD/44516.aspx
ü CP2K: https://www.cp2k.org/about
ü NAMD: http://www.ks.uiuc.edu/Research/namd/
ü GROMACS: http://www.gromacs.org/
ü ….
Flow of
q Commercial software: water
ü Amber: http://ambermd.org/ and
ions
q Home made codes: q Visualization: thru a
ü C, C++ Ø VMD silica
ü Fortran, … etc. pore
Ø OVITO, …
ü Python, … etc. q Analysis?
Summer School, June 25-28, 2018
 Molecular Dynamics: some results

q Binary Metallic alloys:

Ø Melting and crystallization.

Ø Solid-Liquid interfaces.
Ø Crystal growth from melt.
Ø Crystal growth is diffusion limited process.

q Glasses:

Ø Shear deformations in amorphous materials

Ø How to prepare a glass using MD simulation?
Ø Glass Indentation using MD.

Summer School, June 25-28, 2018

 Solid-Liquid interface velocities
Why B2-Al50Ni50?
ü B2-Al50Ni50: prototype of binary ordered metals
ü simulations of interfacial growth in binary systems rare
ü growth kinetics of binary metals: diffusion limited?
ü crystal growth slower than in one-component metals
ü understand crystal growth of alloys on microscopic level

Questions:
Ø crystal growth & accurate estimation of Tm? Ø Frenkel J., Phys. Z.
Ø solid-liquid interface velocity from interface motion? Sowjetunion, 1 (1932)
Ø kinetic coefficients and their anisotropy? 498.
Ø solid-liquid interface motion controlled by mass diffusion? Ø Wilson H.A., Philos.
Mag. , 50 (1900) 238.
Ø solid-liquid coexistence, interface structure?
Ø how to distinguish between solid-like & liquid-like particles?

Summer School, June 25-28, 2018

 Solid-Liquid interfaces: Al-Ni
q solve Newton's equation of motion for system of N particles:
• velocity Verlet algorithm (time step = 1 fs)
• NPT ensemble: Allen M.P. and Tildeslay D.J.,
Ø constant pressure (Anderson algorithm): p = 0 Computer simulation of liquids, 1987
Ø constant temperature: stochastic heat bath Anderson H.C., JCP 72 (1980) 2384
• periodic boundary conditions in all directions
MD of pure systems MD of inhomogeneous systems

Ø lattice properties Ø accurate melting temperature Tm

Ø T dependence of density Ø kinetic coefficients & their anisotropy
Ø Structural quantities Ø solid-melt interface structure
Ø Self-diffusion constant Ø crystal growth

Summer School, June 25-28, 2018

 Simulation parameters
q Binary metallic mixtures - simple: Lennard-Jones potential
- better: EAM 1
q EAM potential: U pot = å u (rkl ) + å F ( r k )
2 k ,l k
Ø two body interactions.
Ø many body interactions (e-density).
Ø fitting to both experimental and ab-initio data.
å
r k = rl (rkl )
l ¹k
Ø reproduces the lattice properties & point defects.
Ø structure and dynamics of AlNi melts. Y. Mishin et al., PRB 65, (2002) 224114.
J. Horbach et al., PRB 75, (2007) 174304.

Ø Solid and liquid properties:

2000 particles (Lx = Ly = Lz = 24.6 Å)

Ø Solid-liquid interfaces (N particles):

NAl = Nni ⇒ D = Lz ≃ 3 × Lx ≃ 3 × Ly
10386 and 12672 particles, … or more
D ≃ 3 × Lx

Summer School, June 25-28, 2018

 Heating and cooling: pure phases
q How to go from crystal to melt & from melt to crystal?
ü start from B2 phase: equilibration at 1000 K
ü try to melt the crystal: heating process
ü cool down the melt: cooling process

Tm = ?
Ø binary alloys: glass formers.
Ø crystallization: process too slow
Ø brute force method:
q How to study crystallization? not appropriate to estimate TM

Summer School, June 25-28, 2018

 Estimation of melting temperature
q How to prepare a system with two pahes?
Ø Equilibrate a crystal (NPT, p=0)
Ø Fix the particles in the middle of the box
Ø Heat away the two other regions
Ø Quench at the target temperature
Estimation of the melting temperature TM from solid-liquid interface motion

T >TM :
Melting

T =TM :
coexistence

T <TM :
crystallization

A. Kerrache et al., EPL 2008.

Interface velocity Enthalpy as a function of time

Summer School, June 25-28, 2018

 Characterization of SL interfaces
Bond order parameter profile Partial particle density profile

v constant density in the liquid region.

å cos(nq xy (i, j, k ))
1
qn =
N i, j ,k
v solid-liquid interface over several layers.
n = 1,2,....,6 v pronounced chemical ordering in the
solid region.
i,j and k: indices for nearest neighbors, q(i,j,k): v mass transport required for crystal
bond angle formed by i, j and k atoms. growth.

Summer School, June 25-28, 2018

 Melting of AlNI at 1600 K

Particle density along the solid-liquid interface

q Melting

Summer School, June 25-28, 2018

 Solid-Liquid interface coexistence at 1518 K
Particle density along the solid-liquid interface

q Coexistence
Summer School, June 25-28, 2018
 Crystallization of AlNi at 1400 K

Particle density along the solid-liquid interface

q Crystallization

Summer School, June 25-28, 2018

 Crystal growth: diffusion limited
q Why the solid-liquid interface
velocity presents a maximum?

ü Maximum of 0.15 m/s at 180 K

Interface velocity divided by the
average self diffusion constant.
ü Maximum due to decreasing
of diffusion constant.
ü Linear regime only up to 30 K
of under-cooling.

Solid-liquid interface velocity as a What about the mass transport

function of temperature across the solid-liquid interface?
Inset: as a function of under-cooling

Summer School, June 25-28, 2018

 Mass transport across the interface
q Role of the mass transport on the
crystal growth:
Ø order parameter to distinguish solid
and liquid particles locally.
Ø compute the particle density and
mass density profiles.
Ø order parameter profile.
Ø number of solid-like particles.
Ø solid-liquid interface velocities from
the number of solid-like particles.
Ø diffusion across the interface.

Mass transport on the liquid phase and across the interface

Summer School, June 25-28, 2018

 Mass transport across the interface
1
Ns
(zis (t) - zis (0))2
Dzs ,a (zs ) = lim
t ®¥ N
å 2t
s is =1

The diffusion constants decrease

when we cross the solid-liquid
interface.

Wilson-Frenkel theory:
activated process controlled by
mass diffusion in the liquid phase

Mass transport and particle density across Wilson H.A. Philos. Mag. , 50 (1900) 238.
Frenkel J., Phys. Z. Sowjetunion, 1 (1932) 498.
the solid-liquid interface A. Kerrache et al. EPL, 2008.
Crystal growth: controlled by mass transport
in the liquid phase and solid-liquid interface Experimental data?

Summer School, June 25-28, 2018

 Comparison to experimental data
Interface velocities: simulation, lab. exp., micro-gravity experiments
ü terrestrial data (Assadi et al.)
ü µg data (parabolic flight) , H.
Hartmann (PhD thesis)
H. Assadi, et al., Acta Mat. 54, 2793 (2006).

good agreement with experimental data

A. Kerrache et al., EPL 81 (2008) 58001.

Summer School, June 25-28, 2018

 Molecular Dynamics: some results

q Binary Metallic alloys:

Ø Melting and crystallization.

Ø Solid-Liquid interfaces.
Ø Crystal growth from melt.
Ø Crystal growth is diffusion limited process.

q Glasses:

Ø Shear deformations in amorphous materials

Ø How to prepare a glass using MD simulation?
Ø Glass Indentation using MD.

Summer School, June 25-28, 2018

 Shear deformations in amorphous silicon
Ø equilibration of the sample at the upper wall
desired temperature (Bulk simulation).
Ø define the lower & upper walls.
Ø move the particles of upper wall

bulk particles
with a fixed shear velocity vs
Ø integrate the equation of motion of
the mobile particles.
Ø equilibration for 5 ns: fixed walls
Ø periodic boundary conditions fixed
in y direction.
q Temperature: lower wall
rescaling the velocities using the Typical starting configuration of s-Si
y and z components

Summer School, June 25-28, 2018

 Shear deformations at 300 K
Potential energy difference ΔE as a
function of imposed strain at 300 K

v Shear velocity: 10-5 to 8x10-2 Å/ps

v Quadratic increase of Ep and strain 10-3 Å/ps è 8 ns
v independent at small strain. 10-5 Å/ps è 800 ns
v Emax increases with shear velocity.
A. Kerrache et al. PRB 83 (2011) 134122.

Summer School, June 25-28, 2018

 Shear deformations at 300 K
Fraction of perfect 4- fold atoms Fraction of 5- fold atoms

q Increasing shear velocity leads:

Ø an increase of the fraction of 5-fold atoms q What will happened if the
Ø a decrease of the fraction of 4-fold atoms shear deformations were
v Increases the disorder and the defects applied at high temperature?

Summer School, June 25-28, 2018

 Crystallization induced by shear

Ø Temperatures: 300, 600,

900 and 1000 K
Ø Shear velocity: 10-5 Å/ps
Ø Total displacement: δ = 5 Å

Potential energy ΔEP as a

function of imposed strain

T 300 600 900 1000

ΔE +0.02 +0.005 -0.01 -0.065
q Increase of the disorder at 300 and 600 K. P

q At 900 K: increase of the order without crystallization. A. Kerrache et al.

q At 1000 K: shear induce the crystallization of a-Si. PRB 84 (2011) 041110.

Summer School, June 25-28, 2018

 Crystallization induced by shear
P. Beaucage et al. PRB (2005).
Coordination
Number

300 K 600 K

Number of
crystalline
particles

900 K 1000 K

Summer School, June 25-28, 2018

 Molecular Dynamics: some results

q Binary Metallic alloys:

Ø Melting and crystallization.

Ø Solid-Liquid interfaces.
Ø Crystal growth from melt.
Ø Crystal growth is diffusion limited process.

q Glasses:

Ø Shear deformations in amorphous materials

Ø How to prepare a glass using MD simulation?
Ø Glass Indentation using MD.

Summer School, June 25-28, 2018

 How to prepare a glass? MD/experiment
Glass preparation diagram:
Cooling rates: 1012 to 1013 K/s

Glass preparation procedure: Model

Ø MD Simulations (DL-POLY).
ü Random configuration (N atoms). Ø System of N particules.
ü Liquid equilibration du at 5000 K (NVT). Ø Time step: 1 fs
ü Cooling per steps of 100 K– (NVT). SBN glasses
ü Glass equilibration at 300 K (NPT). Ø SiO2- B2O3- Na2O
ü Trajectory simulation at 300 K (NVE). ü R = [Na2O] / [B2O3]
ü K = [SiO2] / [B2O3]

Summer School, June 25-28, 2018

 Glass indentation
Movie provided by: Dimitrios Kilymis
Laboratoire Charles Coulomb (L2C), UMR
5221 CNRS-Univ. Montpellier, France. Indenter
Free atoms

Fixed layer

Ø N = 2.1 x 106 atoms

Ø Temperature : 300 K
Ø Speed: 10 m/s
Glass indentation Ø Depth: ~3.0 nm

Summer School, June 25-28, 2018

 Acknowledgments
Prof. Dr. Jürgen Horbach, Dusseldorf, Germany.
Prof. Dr. Kurt Binder, Mainz, Germany.
Prof. A. Meyer and Prof. D. Herlach (DLR), Koln.

Prof. Normand Mousseau, Qc, Canada.

Prof. Laurent J. Lewis, Qc, Canada.

Dr. Dimitrios Kilymis, Montpellier, France.

Prof. Jean-Marc Delaye, CEA, France.

Dr. Victor Teboul, Angers, France.

Prof. Hamid Bouzar, UMMTO, Tizi-Ouzou, Algeria.

Summer School, June 25-28, 2018

 Introduction to LAMMPS

Setting and Running MD

simulations
LAMMPS

q LAMMPS: Molecular Dynamics Simulator (introduction).

q Building LAMMPS step by step.
q Running LAMMPS (input, output, …).
q Benchmark and performance tests.

Summer School, June 25-28, 2018

 Introduction to LAMMPS

LAMMPS
Large-scale Atomic / Molecular Massively
Parallel Simulator

Source: some material and images were adapted from LAMMPS home page

Summer School, June 25-28, 2018

 Where to start to learn LAMMPS?
Large-scale Atomic / Molecular Massively Parallel Simulator
S. Plimpton, A. Thompson, R. Shan, S. Moore, A. Kohlmeyer, …
Sandia National Labs: http://www.sandia.gov/index.html
Ø Home Page: http://lammps.sandia.gov/
Results:
Ø Papers: http://lammps.sandia.gov/papers.html
Ø Pictures: http://lammps.sandia.gov/pictures.html
Ø Movies: http://lammps.sandia.gov/movies.html
Resources:
Ø Online manual: http://lammps.sandia.gov/doc/Manual.html
Ø Search the mailing list: http://lammps.sandia.gov/mail.html
Ø Subscribe to the Mailing List:
https://sourceforge.net/p/lammps/mailman/lammps-users/

Summer School, June 25-28, 2018

 Where LAMMPS has been used?
Ø Biophysics Ø Solid Mechanics

Ø Granular Flow
Ø Material Science
Ø Chemistry

Summer School, June 25-28, 2018

 LAMMPS project page
Big Picture Code Documentation Results Related Tools Context User Support
Pre/Post
Features Download Manual Publications Authors Mail list
processing
Non-features SourceForge Developer guide Pictures Pizza.py Toolkit History Workshops

Latest features & Offsite LAMMPS User scripts

FAQ Tutorials Movies Funding
bug fixes packages & tools and HowTos

MD to LAMMPS Contribute to
Wish list Unfixed bugs Benchmarks Visualization Open source
glossary LAMMPS

Recent LAMMPS News

• (5/18) New fix bond/react command to enable simulation of one or more complex heuristic reactions
that rearrange molecular topology. See details here.
• (3/18) New stable release, 16Mar18 version.
• (9/17) Wrapper on the LATTE DFTB (density-functional tight-binding) quantum code via the fix latte
command. See details here.
• (9/17) USER-MESO package from the Karniadakis group at Brown University, with various
dissipative particle dynamics (DPD) models, including eDPD, mDPD, tDPD. See details here.
• (8/17) New stable release, 11Aug17 version.

Summer School, June 25-28, 2018

 How LAMMPS is designed?
v License
Ø LAMMPS is provided through GNU Public License
https://www.gnu.org/licenses/licenses.en.html#GPL
Ø Free to Use, Modify, and Distribute.
Ø Contribute to LAMMPS: http://lammps.sandia.gov/contribute.html

q Code Layout
Ø C++ and Object-Oriented approach
Ø Parallelization via MPI and OpenMP; runs on GPU.
Ø is invoked by commands through input scripts.
Ø possibility to customized output.
Ø could be interfaced with other codes (python, …): library.
Ø possibility to contribute to LAMMPS: potential, fixes, …

Summer School, June 25-28, 2018

 How to obtain LAMMPS?
v Download Page:
http://lammps.sandia.gov/download.html

Ø Distributions:
Build LAMMPS
üDownload a tarball
from source
üGit checkout and update
üSVN checkout and update
üPre-built Ubuntu executables
üPre-built binary RPMs for Fedora/RedHat/CentOS/openSUSE
üPre-built Gentoo executable
üOS X with Homebrew
üWindows installer package
üApplying patches

Summer School, June 25-28, 2018

 Building LAMMPS
Ø Build from RPMs
üPre-built Ubuntu executables
üPre-built binary RPMs for Fedora/RedHat/CentOS/openSUSE
üPre-built Gentoo executable
Ø Mac does not include all
üOS X with Homebrew packages
super-user access
Ø Install under windows
üWindows installer package

Ø Build from source code

For a customized
üDownload a tarball
installation: build
üGit checkout and update
from source files
üSVN checkout and update Different versions
üApplying patches

Summer School, June 25-28, 2018

 Windows installer for LAMMPS
Ø Download Page: http://rpm.lammps.org/windows.html
Ø Installer: lammps-64bit-latest.exe
Directory:
Program Files\LAMMPS 64-bit 20171023
Binaries under bin: abf_integrate.exe ffmpeg.exe
lmp_mpi.exe restart2data.exe binary2txt.exe chain.exe
lmp_serial.exe msi2lmp.exe createatoms.exe

Ø Execute: lmp_serial.exe < in.lammps

Summer School, June 25-28, 2018
 Building LAMMPS from source
Download the source files:
http://lammps.sandia.gov/download.html#tar

Old version
Current version is: 13 Mar 2018

Archive: lammps-stable.tar.gz
Alternatively, use wget from your terminal:
Ø wget http://lammps.sandia.gov/tars/lammps-stable.tar.gz
Ø wget http://lammps.sandia.gov/tars/lammps-11May18.tar.gz

Summer School, June 25-28, 2018

 LAMMPS source overview
Ø Download and unpack the source code: lammps-stable.tar.gz
Ø LAMMPS directory: lammps-11May18 (lammps-version)
ü bench: benchmark tests (potential, input and output files).
ü doc: documentation (PDF and HTML)
ü examples: input and output files for some simulations
ü lib: libraries to build before building LAMMPS
ü LICENSE and README files.
ü potentials: some of the force fields supported by LAMMPS
ü python: to invoke LAMMPS library from Python
ü src: source files (*.cpp, PACKAGES, USER-PACKAGES, …)
ü tools: some tools like xmovie (similar to VMD but only 2D).

Summer School, June 25-28, 2018

 Building programs
q Common problems:
Ø command not found, ….
Ø undefined reference to fftw, boost, petsc, …
Ø permission denied, …

q Configuration:
ü ./configure --prefix=/home/$USER/software … options
ü cmake .. --DCMAKE_INSTALL_PREFIX=/home/$USER/software
ü ./setup or other provided scripts

q Compile or build the program:

Ø make or make {options}

q Installation:
ü make install

Summer School, June 25-28, 2018

 Building LAMMPS using GNU make
q First: Build libraries if required (atc, meam, reax, …).
q Choose a Makefile: compatible with your system (Compiler, …)
q Choose and install the packages you need.
ü make package # list available packages
ü make package-status (ps) # status of all packages
ü make yes-package # install a single package in src
ü make no-package # remove a single package from src
ü make yes-all # install all packages in src
ü make no-all # remove all packages from src
ü make yes-standard (yes-std) # install all standard packages
ü make no-standard (no-std) # remove all standard packages
ü make yes-user # install all user packages
ü make no-user # remove all user packages
q Build LAMMPS:
Ø make machine # build LAMMPS for machine

Summer School, June 25-28, 2018

 Choose an appropriate Makefile
q machine is one of these from src/MAKE:
Ø # mpi = MPI with its default compiler
Ø # serial = GNU g++ compiler, no MPI
q ... or one of these from src/MAKE/OPTIONS:
Ø # icc_openmpi = OpenMPI with compiler set to Intel icc
Ø # icc_openmpi_link = Intel icc compiler, link to OpenMPI
Ø # icc_serial = Intel icc compiler, no MPI
q ... or one of these from src/MAKE/MACHINES:
Ø # cygwin = Windows Cygwin, mpicxx, MPICH, FFTW
Ø # mac = Apple PowerBook G4 laptop, c++, no MPI, FFTW 2.1.5
Ø # mac_mpi = Apple laptop, MacPorts Open MPI 1.4.3, …
Ø # ubuntu = Ubuntu Linux box, g++, openmpi, FFTW3
q ... or one of these from src/MAKE/MINE: (write your own Makefile)

Summer School, June 25-28, 2018

 Building LAMMPS: demonstration
q Download the latest stable version from LAMMPS home page.
q Untar the archive: tar -xvf lammps-stable.tar.gz
q Change the directory and list the files: cd lammps-11May2018
Ø bench doc examples lib LICENSE potentials python README src tools
q Choose a Makefile (for example: machine=icc_openmpi)
Ø src/MAKE/OPTIONS/Makefile.icc_openmpi
q Load the required modules (Intel, OpenMPI, …)
q Check the packages: package, package-status, yes-package, …
q To build LAMMPS, run: make icc_openmpi
q Add or remove a package (if necessary), then recompile.
q If necessary, edit Makefile and fix the path to libraries.

q Exercise: use an interactive job asking for 4 cores

Ø Compile LAMMPS without any package included.
Ø Add a package and recompile.
Summer School, June 25-28, 2018
 Building LAMMPS: demonstration
q Steps to follow:

1. Download and unpack the source files:

wget http://lammps.sandia.gov/tars/lammps-stable.tar.gz
tar –xvf lammps-stable.tar.gz
2. Submit an interactive job: 4 cores, 1 hour, mem-per-cpu=3500M
3. Load the required modules (eigen, voro++, hdf5, …)
4. Choose and edit the appropriate Makefile: Makefile.icc_openmpi
5. Remove all the packages: make no-all
6. Compile LAMMPS: make icc_openmpi
7. Add one or two packages: make yes-asphere; make yes-voro++
8. Make the necessary changes: paths, libraries, … etc.
9. Clean and recompile the code: make clean-all && make icc_openmpi
10. To add more packages: repeat 7 to 9 (with different packages).

Summer School, June 25-28, 2018

 Building LAMMPS: demonstration
1. wget http://lammps.sandia.gov/tars/lammps-stable.tar.gz

tar –xvf lammps-stable.tar.gz

2. cp -r lammps-16Mar18 build-lammps-16Mar18
3. cd build-lammps-16Mar18/src && make clean-all
4. source ../../get_lammps_dependencies.sh
5. make icc_openmpi
6. make clean-all
7. make yes-voro++
8. Change the Makefile: Makefile.icc_openmpi
9. Recompile: make icc_openmpi

Summer School, June 25-28, 2018

 Command line for running LAMMPS
q Executable: lmp_machine
q Files:
Ø Input File: in.lmp_file
Ø Potential: see examples and last slides for more details.
Ø Initial configuration: can be generated by LAMMPS, or another
program or home made program.
q Interactive Execution:
$ ./lmp_machine < in.lmp_file
$ ./lmp_machine –in in.lmp_file
q Redirect output to a file:
$ ./lmp_machine < in.lmp_file > output_file
$ ./lmp_machine –in in.lmp_file –l output_file

Summer School, June 25-28, 2018

 LAMMPS: modules
v To know more about the modules installed, use "module spider" .
v Search for modules with the name "lammps"
module spider lammps
v Search for all modules that have the character "lammps" in their
names:
module -r spider '.*lammps.*'
v Search of a particular module of interest: lammps-omp/20170811
module spider lammps-omp/20170811
v Load the module of interest: lammps-omp/20170811
module load nixpkgs/16.09 intel/2016.4 openmpi/2.1.1 lammps-
omp/20170811
v Check if the module is correctly loaded: module list
v For more information:
module show lammps-omp/20170811

Summer School, June 25-28, 2018

 Command line options
q Command-line options:
At run time, LAMMPS recognizes several optional command-line switches
which may be used in any order.
-e or -echo, -h or –help, -i or –in, -k or –kokkos, -l or –log, -nc or –nocite,
-pk or –package, -p or –partition, -pl or –plog, -ps or –pscreen, -r or –
restart, -ro or –reorder, -sc or –screen, -sf or –suffix, -v or –var

q As an interactive job:
Ø mpirun -np 16 lmp_machine -in in.alloy
Ø mpiexec -n 4 lmp_machine < in.alloy
q As a submitted job (Torque, SLURM, …):
Ø mpiexec lmp_machine < in.alloy > my.log
Ø mpirun lmp_machine < in.alloy > my.log
Ø srun lmp_machine < in.alloy > my.log

Summer School, June 25-28, 2018

 Simulation run overview

INPUT RUNNING OUTPUT

• Initial positions • Trajectories
• Initial velocities • Velocities
• Time step • Molecular • Forces
• Mass Dynamics • Energy
• Charges Simulation (NPT, • Temperature
NVT, NVE)
• PBC (or fixed) • Pressure
• Minimization
• Units • Density
• Monte Carlo
• Potential • Snapshots
• Atomic to
• Ensemble • Movies
Continuum
• …. etc. • … etc.

Summer School, June 25-28, 2018

 Overview of a simulation settings
q Command Line:
Ø Every simulation is executed by supplying an input text script to the
LAMMPS executable: lmp < lammps.in > log_lammps.txt
q Parts of an input script:
Ø Initialize: units, dimensions, PBC, etc.
Ø Atomic positions (built in or read from a file) and velocities.
Ø Settings:
ü Inter-atomic potential (pair_style, pair_coeff)
ü Run time simulation parameters (e.g. time step)
ü Fixes: operations during dynamics (e.g. thermostat)
ü Computes: calculation of properties during dynamics
ü Rendering: snapshots of MD trajectory, movie.
q Run the simulation for N steps (time step = dt).

Summer School, June 25-28, 2018

 LAMMPS input file example: LJ melt
# 3d Lennard-Jones melt Comment

units lj
Define units
atom_style atomic

lattice fcc 0.8442 Create the simulation box

region box block 0 10 0 10 0 10
Or read data from a file
create_box 1 box
create_atoms 1 box
mass 1 1.0
Initialize the
velocity all create 3.0 87287
velocities
# Potential
pair_style lj/cut 2.5 Define the
pair_coeff 1 1 1.0 1.0 2.5 potential

Summer School, June 25-28, 2018

 LAMMPS input file example: LJ melt
# Neighbour list: Monitor the
neighbor 0.3 bin
neighbour list
neigh_modify every 20 delay 0 check no
Thermodynamic
# set the thermodynamic ensemble:
fix 1 all nve Ensemble

Store the
dump id all atom 50 dump.melt
#dump_modify …. trajectory

log log.melt
Log file:
thermo_style custom step temp etotal ….
customize output
thermo 50
Run the simulation
run 250
for N steps
# End of the simulation.

Summer School, June 25-28, 2018

 LAMMPS: input commands
q Initialization

Ø Parameters: set parameters that need to be defined before atoms are created:
units, dimension, newton, processors, boundary, atom_style, atom_modify.

Ø If force-field parameters appear in the files that will be read:

pair_style, bond_style, angle_style, dihedral_style, improper_style.

Ø Atom definition: there are 3 ways to define atoms in LAMMPS.

ü Read them in from a data or restart file via the read_data or read_restart
commands.
ü Or create atoms on a lattice (with no molecular topology), using these
commands: lattice, region, create_box, create_atoms.
ü Duplicate the box to make a larger one the replicate command.

Summer School, June 25-28, 2018

 LAMMPS: settings
q Once atoms are defined, a variety of settings need to be specified:
force field coefficients, simulation parameters, output options …
v Force field coefficients:
pair_coeff, bond_coeff, angle_coeff, dihedral_coeff, improper_coeff,
kspace_style, dielectric, special_bonds.
v Various simulation parameters:
neighbor, neigh_modify, group, timestep, reset_timestep, run_style,
min_style, min_modify.
v Fixes: nvt, npt, nve, …
v Computations during a simulation:
compute, compute_modify, and variable commands.
v Output options: thermo, dump, and restart commands.

Summer School, June 25-28, 2018

 Customize output
thermo freq_steps Thermodynamic
thermo_style style args properties
Ø style = one or multi or custom
Ø args = list of arguments for a particular style
one args = none
multi args = none custom
args = list of keywords possible
q keywords = step, elapsed, elaplong, dt, time, cpu, tpcpu, spcpu, cpuremain,
part, timeremain, atoms, temp, press, pe, ke, etotal, enthalpy, evdwl, ecoul,
epair, ebond, eangle, edihed, eimp, emol, elong, etail, vol, density, lx, ly, lz, xlo,
xhi, ylo, yhi, zlo, zhi, xy, xz, yz, xlat, ylat, zlat, bonds, angles, dihedrals,
impropers, pxx, pyy, pzz, pxy, pxz, pyz ….. Etc

q Example:
thermo_style custom step temp press pe ke etotal density lx ly lz

Summer School, June 25-28, 2018

 Customize output
v dump command: Trajectories
Options: vtk, h5md, molfile, netcdf, image, movie
Snapshots Movie
v Syntax:
dump ID group-ID style N file args
Ø D = user-assigned name for the dump
Ø group-ID = ID of the group of atoms to be dumped
Ø style = atom or atom/gz or atom/mpiio or cfg or cfg/gz or cfg/mpiio or custom
or custom/gz or custom/mpiio or dcd or h5md or image or or local or molfile or
movie or netcdf or netcdf/mpiio or vtk or xtc or xyz or xyz/gz or xyz/mpiio
Ø N = dump every this many time steps
Ø file = name of file to write dump info to
Ø args = list of arguments for a particular style

v Example:
Ø dump myDump all atom 100 dump.atom
Ø dump 2 subgroup atom 50 dump.run.bin

Summer School, June 25-28, 2018

 Running LAMMPS: demonstration
q After compiling LAMMPS, run some examples:

q Where to start to learn LAMMPS?

Ø Make a copy of the directory examples to your working directory.
Ø Choose and example to run.
Ø Indicate the right path to the executable or use modules available (if any).
Ø Edit the input file and check all the parameters.
Ø Check the documentation for the commands and their arguments.
Ø Run the test case: lmp_icc_openmpi < in.melt
Ø Check the output files (log files), plot the thermodynamic properties, ...
Ø Use VMD (or any other software) for visualization.

Summer School, June 25-28, 2018

 Running LAMMPS: demonstration
q Connect to cedar and/or graham
ssh –Y [email protected]
ssh –Y [email protected]
q Go to the directory where you copied or download the exercises.
q To run LAMMPS interactively, submit an inteactive job using salloc
salloc –ntasks=1 –mem-per-cpu=2500M –time=00-00:30
q Submit some jobs using sbatch: sbatch your_script.sh
q Edit the input files
q Run the jobs
q Check the output files.

Summer School, June 25-28, 2018

 Output example: LJ melt
LAMMPS (30 Jul 2016)
using 1 2 OpenMP thread(s) per MPI task
# 3d Lennard-Jones melt
units ljatom_style atomic
lattice fcc 0.8442Lattice spacing in x,y,z = 1.6796 1.6796 1.6796
region box block 0 10 0 10 0 10
create_box 1 box
Created orthogonal box = (0 0 0) to (16.796 16.796 16.796) 2 by 2 by 3 MPI
processor grid
create_atoms 1 box
Created 4000 atoms
mass 1 1.0

Summer School, June 25-28, 2018

 Output example: LJ melt
thermo 100
run 25000
Neighbor list info ...
1 neighbor list requests
update every 20 steps, delay 0 steps, check no
max neighbors/atom: 2000, page size: 100000
master list distance cutoff = 2.8
ghost atom cutoff = 2.8
binsize = 1.4 -> bins = 12 12 12
Memory usage per processor = 2.05293 Mbytes
Step Temp E_pair E_mol TotEng Press
0 3 -6.7733681 0 -2.2744931 -3.7033504
100 1.6510577 -4.7567887 0 -2.2808214 5.8208747
200 1.6393075 -4.7404901 0 -2.2821436 5.9139187
300 1.6626896 -4.7751761 0 -2.2817652 5.756386

Summer School, June 25-28, 2018

 Output example: LJ melt
25000 1.552843 -4.7611011 0 -2.432419 5.7187477
Loop time of 15.4965 on 12 procs for 25000 steps with 4000 atoms
Performance: 696931.853 tau/day, 1613.268 timesteps/s
90.2% CPU use with 12 MPI tasks x 1 OpenMP threads
MPI task timing breakdown:
Section | min time | avg time | max time |%varavg| %total
-----------------------------------------------------------------------------
Pair | 6.6964 | 7.1974 | 7.9599 | 14.8 | 46.45
Neigh | 0.94857 | 1.0047 | 1.0788 | 4.3 | 6.48
Comm | 6.0595 | 6.8957 | 7.4611 | 17.1 | 44.50
Output | 0.01517 | 0.01589 | 0.019863 | 1.0 | 0.10
Modify | 0.14023 | 0.14968 | 0.16127 | 1.7 | 0.97
Other | | 0.2332 | | | 1.50
Total wall time: 0:00:15

Summer School, June 25-28, 2018

 Benchmarking: potential

1. granular 13. spce Parameters:

2. fene 14. protein
3. lj 15. gb Ø 24 different cases.
4. dpd 16. reax_AB Ø Number of particles:
5. eam 17. airebo about 32000
6. sw 18. reaxc_rdx Ø CPUs = 1
7. rebo 19. smtbq_Al Ø MD steps = 1000
8. tersoff 20. vashishta_table_sio2 Ø Record the simulation
9. eim 21. eff time and the time used
10. adp 22. comb in computing the
11. meam 23. vashishta_sio2 interactions between
12. peri 24. smtbq_Al2O3 particles.

Summer School, June 25-28, 2018

 Benchmarking: potential

Summer School, June 25-28, 2018

 Performance test: LJ
q Directory: Benchmark

q Simulation: LJ Melt

q Number of atoms: 2048, 4000, 6912, 13500

q Exercise:

Ø Submit the jobs using different number of cores: 1,2, 4, 8, 16

Ø For each system: collect the data:

Ø time used for pair interactions as a function of number of cores.

Ø time used for communications as a function of number of cores.

Summer School, June 25-28, 2018

 Performance test: LJ
Time spend for pair interactions computing and communications as a
function of number of cores for different systems

Cores Pairs Comm Pairs Comm Pairs Comm Pairs Comm

1 73.68 1.36 73.70 1.28 73.66 1.27 73.72 1.29

2 70.35 5.19 70.77 4.68 70.51 5.11 67.80 8.77

4 62.77 13.98 64.93 12.19 67.52 8.99 67.74 8.71

8 58.36 20.14 61.78 15.58 64.10 12.86 62.06 8.71

16 56.69 20.18 56.70 20.18 56.97 19.80 56.41 20.38

2048 4000 6912 13500

Summer School, June 25-28, 2018

 Performance test: Tersoff potential

CPU time used for computing the interactions between particles

as a function the number of processors for different system size.

Summer School, June 25-28, 2018

 Performance test: Tersoff potential

CPU time used for computing the interactions between particles

as a function the number of processors for different system size.
Summer School, June 25-28, 2018
 Performance test: Tersoff potential
q Domain decomposition

Ø Size, shape of the system.

Ø Number of processors.
Ø size of the small units.
Ø correlation between the
communications and the
number of small units.
Ø Reduce the number of cells
to reduce communications.

Summer School, June 25-28, 2018

 Learn more about LAMMPS
q Home Page: http://lammps.sandia.gov/
q Examples: deposit, friction, micelle, obstacle, qeq, streitz, MC, body, dipole,
hugoniostat, min, peptide, reax, tad, DIFFUSE, colloid, indent, msst, peri,
rigid , vashishta, ELASTIC, USER, comb, eim, nb3b, pour, shear, voronoi,
ELASTIC_T, VISCOSITY, coreshell, ellipse, meam, neb, prd, snap, HEAT,
accelerate, crack, flow, melt, nemd
q Results: Ø Papers: http://lammps.sandia.gov/papers.html
Ø Pictures: http://lammps.sandia.gov/pictures.html
Ø Movies: http://lammps.sandia.gov/movies.html

q Resources:
Ø Online Manual: http://lammps.sandia.gov/doc/Manual.html
Ø Search the mailing list: http://lammps.sandia.gov/mail.html
Ø Mailing List:
https://sourceforge.net/p/lammps/mailman/lammps-users/

Summer School, June 25-28, 2018

 Potentials: classified by materials

Ø Bio-molecules: CHARMM, AMBER, OPLS, COMPASS (class 2),

long-range Coulombic via PPPM, point dipoles, ...
Ø Polymers: all-atom, united-atom, coarse-grain (bead-spring FENE),
bond-breaking, …
Ø Materials: EAM and MEAM for metals, Buckingham, Morse, Yukawa,
Stillinger-Weber, Tersoff, EDIP, COMB, SNAP, ...
Ø Chemistry: AI-REBO, REBO, ReaxFF, eFF
Ø Meso-scale: granular, DPD, Gay-Berne, colloidal, peri-dynamics,
DSMC...
Ø Hybrid: combine potentials for hybrid systems: water on metal,
polymers/semiconductor interface, colloids in solution, …

Summer School, June 25-28, 2018

 Potentials: classified by functional form
Ø Pair-wise potentials: Lennard-Jones, Buckingham, ...
Ø Charged Pair-wise Potentials: Coulombic, point-dipole
Ø Many-body Potentials: EAM, Finnis/Sinclair, modified EAM
(MEAM), embedded ion (EIM), Stillinger-Weber, Tersoff, AI-REBO,
ReaxFF, COMB
Ø Coarse-Grained Potentials: DPD, GayBerne, ...
Ø Meso-scopic Potentials: granular, peri-dynamics
Ø Long-Range Electrostatics: Ewald, PPPM, MSM
Ø Implicit Solvent Potentials: hydrodynamic lubrication, Debye
Ø Force-Field Compatibility with common: CHARMM, AMBER, OPLS,
GROMACS options

Summer School, June 25-28, 2018

 Force fields: examples
pair_style none - turn off pairwise interactions
pair_style hybrid - multiple styles of pairwise interactions
pair_style hybrid/overlay - multiple styles of superposed pairwise interactions
pair_style zero - neighbor list but no interactions
pair_style adp - angular dependent potential (ADP) of Mishin
pair_style airebo - AIREBO potential of Stuart
pair_style airebo/morse - AIREBO with Morse instead of LJ
pair_style beck - Beck potential
pair_style body - interactions between body particles
pair_style bop - BOP potential of Pettifor
pair_style born - Born-Mayer-Huggins potential
pair_style born/coul/long - Born-Mayer-Huggins with long-range Coulombics
pair_style born/coul/long/cs - Born-Mayer-Huggins with long-range Coulombics
and core/shell
pair_style born/coul/msm - Born-Mayer-Huggins with long-range MSM
Coulombics

Summer School, June 25-28, 2018

 Force fields: examples
pair_style born/coul/wolf - Born-Mayer-Huggins with Coulombics via Wolf
potential
pair_style brownian - Brownian potential for Fast Lubrication Dynamics
pair_style brownian/poly - Brownian potential for Fast Lubrication Dynamics with
polydispersity
pair_style buck - Buckingham potential
pair_style buck/coul/cut - Buckingham with cutoff Coulomb
pair_style buck/coul/long - Buckingham with long-range Coulombics
pair_style buck/coul/long/cs - Buckingham with long-range Coulombics and
core/shell
pair_style buck/coul/msm - Buckingham long-range MSM Coulombics
pair_style buck/long/coul/long - long-range Buckingham with long-range
Coulombics
pair_style colloid - integrated colloidal potential
pair_style comb - charge-optimized many-body (COMB) potential
pair_style comb3 - charge-optimized many-body (COMB3) potential

Summer School, June 25-28, 2018

 Force fields: examples
pair_style coul/cut - cutoff Coulombic potential
pair_style coul/debye - cutoff Coulombic potential with Debye screening
pair_style coul/dsf - Coulombics via damped shifted forces
pair_style coul/long - long-range Coulombic potential
pair_style coul/long/cs - long-range Coulombic potential and core/shell
pair_style coul/msm - long-range MSM Coulombics
pair_style coul/streitz - Coulombics via Streitz/Mintmire Slater orbitals
pair_style coul/wolf - Coulombics via Wolf potential
pair_style dpd - dissipative particle dynamics (DPD)
pair_style dpd/tstat - DPD thermostatting
pair_style dsmc - Direct Simulation Monte Carlo (DSMC)
pair_style eam - embedded atom method (EAM)
pair_style eam/alloy - alloy EAM
pair_style eam/fs - Finnis-Sinclair EAM
pair_style eim - embedded ion method (EIM)
pair_style gauss - Gaussian potential

Summer School, June 25-28, 2018

 Force fields: examples
pair_style gayberne - Gay-Berne ellipsoidal potential
pair_style gran/hertz/history - granular potential with Hertzian interactions
pair_style gran/hooke - granular potential with history effects
pair_style gran/hooke/history - granular potential without history effects
pair_style hbond/dreiding/lj - DREIDING hydrogen bonding LJ potential
pair_style hbond/dreiding/morse - DREIDING hydrogen bonding Morse potential
pair_style kim - interface to potentials provided by KIM project
pair_style lcbop - long-range bond-order potential (LCBOP)
pair_style line/lj - LJ potential between line segments
pair_style lj/charmm/coul/charmm - CHARMM potential with cutoff Coulomb
pair_style lj/charmm/coul/charmm/implicit - CHARMM for implicit solvent
pair_style lj/charmm/coul/long - CHARMM with long-range Coulomb
pair_style lj/charmm/coul/msm - CHARMM with long-range MSM Coulombics
pair_style lj/class2 - COMPASS (class 2) force field with no Coulomb
pair_style lj/class2/coul/cut - COMPASS with cutoff Coulomb

Summer School, June 25-28, 2018

 Force fields: examples
pair_style lj/gromacs/coul/gromacs - GROMACS-style LJ and Coulombic potential
pair_style lj/long/coul/long - long-range LJ and long-range Coulombics
pair_style lj/long/dipole/long - long-range LJ and long-range point dipoles
pair_style lj/long/tip4p/long - long-range LJ and long-range Coulomb for TIP4P
water
pair_style lj/smooth - smoothed Lennard-Jones potential
pair_style lj/smooth/linear - linear smoothed Lennard-Jones potential
pair_style lj96/cut - Lennard-Jones 9/6 potential
pair_style lubricate - hydrodynamic lubrication forces
pair_style lubricate/poly - hydrodynamic lubrication forces with polydispersity
pair_style lubricateU - hydrodynamic lubrication forces for Fast Lubrication
Dynamics
pair_style lubricateU/poly - hydrodynamic lubrication forces for Fast Lubrication
with polydispersity
pair_style meam - modified embedded atom method (MEAM)

Summer School, June 25-28, 2018

 Force fields: examples
pair_style mie/cut - Mie potential
pair_style morse - Morse potential
pair_style nb3b/harmonic - nonbonded 3-body harmonic potential
pair_style nm/cut - N-M potential
pair_style nm/cut/coul/cut - N-M potential with cutoff Coulomb
pair_style nm/cut/coul/long - N-M potential with long-range Coulombics
pair_style peri/eps - peridynamic EPS potential
pair_style peri/lps - peridynamic LPS potential
pair_style peri/pmb - peridynamic PMB potential
pair_style peri/ves - peridynamic VES potential
pair_style polymorphic - polymorphic 3-body potential
pair_style reax - ReaxFF potential
pair_style rebo - 2nd generation REBO potential of Brenner
pair_style resquared - Everaers RE-Squared ellipsoidal potential
pair_style snap - SNAP quantum-accurate potential
pair_style soft - Soft (cosine) potential

Summer School, June 25-28, 2018

 Force fields: examples
pair_style sw - Stillinger-Weber 3-body potential
pair_style table - tabulated pair potential
pair_style tersoff - Tersoff 3-body potential
pair_style tersoff/mod - modified Tersoff 3-body potential
pair_style tersoff/zbl - Tersoff/ZBL 3-body potential
pair_style tip4p/cut - Coulomb for TIP4P water w/out LJ
pair_style tip4p/long - long-range Coulombics for TIP4P water w/out LJ
pair_style tri/lj - LJ potential between triangles
pair_style vashishta - Vashishta 2-body and 3-body potential
pair_style yukawa - Yukawa potential
pair_style yukawa/colloid - screened Yukawa potential for finite-size particles
pair_style zbl - Ziegler-Biersack-Littmark potential

Summer School, June 25-28, 2018