This project is a dissertation, contributing towards a MEng Software Engineering degree. The full report including
implementation and demonstration fulfils 50% of the third year. The project is split into three parts - the description,
outlining the problem the project is set out to solve. The second part is the survey and analysis report, which critically
studies techniques that set out to fulfil the same goal. The third part is the final report, detailing the precise method
and result of the implementation. This is then followed by a demonstration of the project running.
Computer games are becoming more popular and the graphics processing power is increasing, allowing developers
to create more realistic graphics. Much effort has gone into making efficient and believable simulations of the
cracking process in computer games. Most previous techniques simulate cracking with a set animation or as an
extension to an existing physics engine.
This report concentrates on developing a plausible three dimensional cracking algorithm. The emphasis is on
geometry of the object being cracked, along with the material properties and impact variables, rather than purely
on an artist’s impression or physical simulation. The overall result is a progressive cracking object with
believable cracking detail, which is maintained at real time speeds.