Simulations

The simulation module contains types representing algorithms. The main type is the Simulation type, which is parametrized by a Propagator. This propagator will determine which kind of simulation we are running: Molecular Dynamics; Monte-Carlo; energy minimization; etc.

Note

Only molecular dynamic is implemented in Jumos for now, but at least Monte-Carlo and energetic optimization should follow.

• Molecular Dynamics
  • Time integration
  • Force computation
  • Controlling the simulation
  • Checking the simulation consistency
  • Default algorithms
  • Functions for algorithms selection
• Computing values of interest
  • Energy related values
  • Volume
  • Pressure
• Exporting values of interest
  • Existing outputs
  • Defining a new output

Propagator

The Propagator type is responsible for generating new frames in the simulated universe. If you want to help adding a new propagator to Jumos, please signal yourself in the Gihtub issues list.

Simulation type

In Jumos, simulations are first-class citizen, i.e. objects bound to variables. The following constructors should be used to create a new simulation:

Simulation(propagator::Propagator)

Create a simulation with the specified propagator.

Simulation(propagator::Symbol, args...)

Create a simulation with a propagator which type is given by the propagator symbol. The args are passed to the propagator constructor.

If propagator takes one of the :MD, :md and :moleculardynamics values, a MolecularDynamics propagator is created.

The main function to run the simulation is the propagate!() function.

propagate!(simulation, universe, nsteps)

Propagate an universe for nsteps steps, using the simulation method. Usage example:

julia> sim = Simulation(:MD, 1.0)

julia> universe # This should be an universe, either read from a file or built by hand

julia> propagate!(sim, universe, 4000) # Run the MD simulation for 4000 steps