Welcome to the hands-on course ICASCM 2025#

Course Overview#

This course is designed as a hands-on tutorial that introduces the use of molecular dynamics (MD) simulations in the study of cementitious materials. It provides a practical first overview of how atomistic modelling can be applied to this field.

As a case study, we will focus on a paradigmatic example that has been widely studied in the cement research community: the calculation of ion diffusivity in C-S-H gel nanopores (see Duque-Redondo et al., 2022 for a review). We will:

  • Construct a realistic slit-pore C–S–H model saturated with water and NaCl.

  • Prepare a LAMMPS input file to run molecular dynamics simulations.

  • Analyse the results and calculate:

    • the diffusion constants of the ions inside the pore.

    • the density profiles of the ions inside the pore.

The aim is to illustrate two complementary approaches:

  1. A simple tool-based route, intended for novice users taking their first steps in molecular simulations.

  2. A programming-based route for more experienced users, illustrating how to automate the simulations and analyses. It may initially seem more demanding, but it quickly becomes straightforward and will increase your productivity.

Each section contains detailed explanations of what to do, together with the instructions for the corresponding exercise, and an estimated time to complete it. In case that you do not complete the exercise, a solution will be available in the course folder so that you can continue with the class.

Software Requirements#

To follow the tutorial, participants will need to prepare a few tools in advance:

  • Visualisation programs: We will use Ovito or VESTA. These are free visualization tools, available for Windows, Linux, and macOS:

    • OVITO is excellent for analysing atomistic trajectories and creating high-quality animations, focused on materials science.

    • VESTA is widely used for crystallographic structures and static visualisations.

  • Python execution environment: We will use pyCSH and small Python scripts to post-process simulation data. A minimal understanding of how to navigate file paths and execute Python scripts will be necessary. For editing and running scripts, we like Visual Studio Code (VS Code), but you can run them from your terminal or any other preferred option.

  • Model setup: We will use several codes to set up our model and simulation: VMD, pyCSH, and Packmol (advanced):

    • pyCSH is a code to automatically create hundreds of unique and realistic C-S-H models.

    • VMD is a visualuzation and analysis code designed for biomolecules, but we will use it here for topology recognition.

    • Packmol is a code to efficiently pack molecules in multiple shapes.

  • Simulation code: We will use LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator). LAMMPS is an open-source classical molecular dynamics simulation code focusing on materials modelling, originally developed at the Sandia National Laboratories. Usually, LAMMPS is installed and run on HPC systems, but here we will use Atomify, a lightweight graphical front-end for LAMMPS. Atomify allows you to build, run, and visualise molecular dynamics simulations in real time through an interactive interface.

Note

No access to a supercomputer is required, since the tutorial is based on a basic but representative example designed to be executable on a personal computer.

  • Postprocessing and Analysis: Postprocessing will be done with Python and TRAVIS (Trajectory Analyzer and Visualizer). TRAVIS is a free program that can process MD trajectories to compute properties such as mean square displacement (MSD) and diffusion coefficients, density and radial distribution profiles, vibrational spectra, velocity autocorrelation functions, and more. TRAVIS is available for Windows, Linux, and macOS.

Tip

There are many more codes and tools that can be helpful for performing MD simulations. We always strongly advocate the use of open-source and freely available software, as it ensures transparency, reproducibility, and accessibility for the entire scientific community.

If you get stucked during the course, you can find all the files that you should use as input and all the files that will be generated as intermidiate and final results in this github folder. Use them if necessary.