A multitude of different methods for atomistic simulations is widely used in both the natural and engineering sciences. Interactions between atoms can be described by wavefunction-based ab initio, density-functional theory (DFT) or semiempirical quantum chemical approaches of differing accuracy or by empirical classical potentials known as force fields. Thus, atomistic simulations can describe essentially all aspects of matter, from structure electronic properties, opening the way to numerous central physical, chemical and biological properties of molecular systems and solids. These properties find use both in pure scientific research and in applications. Once interatomic interactions can be described, time evolution on the atomic scale becomes accessible. Most commonly, classical (force-field based) molecular dynamics (MD) are used; often with enhanced-sampling techniques designed to simulate processes on long time scales. These include hyperdynamics, multiple-replica techniques, MD/Monte Carlo, (adaptive) kinetic Monte Carlo ((a)kMC) or diffusive molecular dynamics (DMD). Atomistic simulation methods are also used to investigate static structures using energy minimization, both for equilibrium structures and transition states to give energy barriers, or for chemical equilibria using Monte Carlo (MC) simulations. Atomistic methods can also be combined with coarse-grained models and particle-based systems, or coupled with the Finite Element Method (FEM) or an environment modelled as a continuum.
The FAU is unique in the field of atomistic simulation because it offers the whole spectrum of both methods and fields of application. Scientists at FAU use methods from classical MD to quantum chemistry and everything in between. The methods are applied in chemistry, biology, physics, medicine, materials science and engineering.
The Atomic Structure Simulation Lab is a Germany-wide unique interdisciplinary competence centre that helps users to select and use atomistic simulation methods in an HPC environment and actively accompanies and coordinates the development of high-performance simulation codes. An interdisciplinary approach promises not only synergy effects, e.g. through the exchange and joint development of simulation and evaluation tools, but in particular a cross fertilization of materials and life sciences, which often use the same or very similar simulation techniques.
A further core project is the education and lifelong training of scientists and engineers. The close cooperation between theory, simulation and experiment, which has a long tradition in Erlangen and was honed in structured research programmes such as the Cluster of Excellence Engineering of Advanced Materials (EAM), ensures that the training is not aimed specifically at modellers, but that it is also made available to experimental colleagues. This is of particular importance against the background of increasing digitalisation in science. Here, consulting and service in the field of research-data management will also be offered.
The Atomistic Simulation Lab is making an important contribution to the key technologies of scientific computing and scientific software development through the sustained concentration of methodological competence both in the application and development of computer codes and their hardware-related optimization.