Incremental update of electrostatic interactions in adaptively restrained particle simulations

Area of Science:

• Computational Physics
• Molecular Dynamics Simulations

Background:

• Calculating long-range potentials is computationally intensive in molecular dynamics.
• Existing methods aim to reduce the computational cost of these calculations.
• Adaptively restrained systems, where only some particles move per timestep, present unique challenges.

Purpose of the Study:

• To develop a fast method for computing electrostatic potentials in adaptively restrained molecular dynamics systems.
• To leverage the specific nature of adaptively restrained systems for computational gains.

Main Methods:

• Developed an incremental algorithm based on a multigrid approach, an alternative to Fourier-based methods.
• Implemented the algorithm within the LAMMPS molecular dynamics simulation package.
• Evaluated the method's performance on various simulated systems.

Main Results:

• The new algorithm's computational complexity scales with the number of active particles.
• Demonstrated superior performance compared to the Particle Particle Particle Mesh (P3M) method for adaptively restrained simulations.
• The method is efficient for systems where particle movement is restricted at each step.

Conclusions:

• The proposed incremental, multigrid-based algorithm provides a computationally efficient solution for electrostatic potential calculations in adaptively restrained molecular dynamics.
• This method offers a significant advantage over existing techniques like P3M in specific simulation scenarios.
• The implementation in LAMMPS makes it accessible for broader use in the molecular dynamics community.