Supervisor: Marcel Mudrich

Many questions in the realm of large atomic and molecular systems are largely incalculable, owing to the complexity of the mathematical models describing them. Traditionally the solution to these problems have been to look for ways to simplify the system without removing its essential characteristics. However, with the advent of computers, and the large jumps in calculation power they’ve brought along, other avenues have been opened. Molecular dynamics is one of these computational methods. Large systems of proteins, crystals or liquids can be modelled as atoms interacting through specially chosen potentials. Computers aren’t perfect however, and several concessions must be made to bring these systems into the specialised world of floating point numbers, numerical integration, and limited computing power. In this talk, I’d like to explore what these concession look like and how one can be certain of the accuracy of these calculations.