Using Atomistic-Compatible Packages

There are three main steps to using an implementation of the Atomistic API. This simple toy example will walk you through each step using Molly.jl for the dynamics. Some of the values from this example were taken from this guide.

Step 0: Load Dependencies

using Atomistic
using InteratomicPotentials
using Molly

Step 1A: Configuring the System

First, we must create an AtomsBase.jl-style system. Here we will create a small Argon cluster with periodic boundary conditions. We are using a helper function to initialize the system automatically with a particular temperature.

N = 864
box_size = 3.47786u"nm"
reference_temp = 94.4u"K"

initial_system = generate_atoms_in_cubic_cell(N, :Ar, box_size, reference_temp)

FlexibleSystem(Ar₈₆₄, AtomsBase.Periodic, box=[[3.47786, 0.0, 0.0], [0.0, 3.47786, 0.0], [0.0, 0.0, 3.47786]]u"nm")

Step 1B: Configuring the Simulator

Second, we initialize our simulator. In this case we are using an AndersenThermostat which is provided by Molly.

Δt = 1e-2u"ps"
steps = 2000
thermostat = Molly.AndersenThermostat(reference_temp, Δt * 10)
simulator = MollySimulator(Δt, steps, coupling=thermostat)

MollySimulator{Molly.VelocityVerlet, Quantity{Float64, 𝐓, Unitful.FreeUnits{(ps,), 𝐓, nothing}}, Molly.AndersenThermostat{Quantity{Float64, 𝚯, Unitful.FreeUnits{(K,), 𝚯, nothing}}, Quantity{Float64, 𝐓, Unitful.FreeUnits{(ps,), 𝐓, nothing}}}}(0.01 ps, 2000, 0.0 ps, Molly.AndersenThermostat{Quantity{Float64, 𝚯, Unitful.FreeUnits{(K,), 𝚯, nothing}}, Quantity{Float64, 𝐓, Unitful.FreeUnits{(ps,), 𝐓, nothing}}}(94.4 K, 0.1 ps), 1)

Step 1C: Configuring the Potential

Lastly, we specify the interatomic potential that we will use for the simulation, Lennard-Jones in this case.

potential = InteratomicPotentials.LennardJones(1.657e-21u"J", 0.34u"nm", 0.765u"nm", [:Ar])

InteratomicPotentials.LennardJones{Float64}(0.00038006812452886086, 6.425068823727619, 14.456404853387143, (:Ar,))

Step 2: Running the Simulation

result = simulate(initial_system, simulator, potential)

MollyResult(System with 864 atoms, box size Quantity{Float64, 𝐋, Unitful.FreeUnits{(nm,), 𝐋, nothing}}[3.47786 nm, 3.47786 nm, 3.47786 nm], MollySimulator{Molly.VelocityVerlet, Quantity{Float64, 𝐓, Unitful.FreeUnits{(ps,), 𝐓, nothing}}, Molly.AndersenThermostat{Quantity{Float64, 𝚯, Unitful.FreeUnits{(K,), 𝚯, nothing}}, Quantity{Float64, 𝐓, Unitful.FreeUnits{(ps,), 𝐓, nothing}}}}(0.01 ps, 2000, 0.0 ps, Molly.AndersenThermostat{Quantity{Float64, 𝚯, Unitful.FreeUnits{(K,), 𝚯, nothing}}, Quantity{Float64, 𝐓, Unitful.FreeUnits{(ps,), 𝐓, nothing}}}(94.4 K, 0.1 ps), 1))

Step 3: Analyzing the Results

We can now analyze the simulation results. The Atomistic API exposes a variety of quantities from each timestep of the simulation (more details here). In this example we will look at the temperature, energy, and radial distribution function.

Temperature

plot_temperature(result, simulator.steps ÷ 200)

Energy

plot_energy(result, simulator.steps ÷ 200)

Radial Distribution Function (RDF)

plot_rdf(result, potential.σ, Int(0.95 * steps))

This page was generated using Literate.jl.