Graphite is a key component of most UK operational reactors and for the most exciting designs of new high temperatures reactors that should one day produce the clean fuel, hydrogen. Graphite acts as a moderator to slow neutrons down and make them more effective for nuclear fission. It is also a structural component, so the otherwise slippery and weak single crystal graphite is not used but rather the components are polycrystalline (in the same way that a rock comprises many different interlocking mineral crystallites). In the course of its neutron moderation it becomes damaged, more porous and the individual crystallites change their shape. These changes are carefully monitored but we need to be able to predict the changes so that we can better gauge the life expectancy of our reactors. It will be an important step towards meeting the UK's commitments to carbon emission reduction to 2020 and beyond. In the longer term, High Temperature gas-cooled Reactors (HTRs) are internationally seen as an important source of power, in particular for hydrogen production, so we need similarly to show that future international HTRs could be capable of operating for 60-100 years. Materials Test reactor data for nuclear graphite are incomplete due to the early termination of irradiation experiments aimed at giving lifetime data for UK AGRs.When the original theories of graphite were formulated in the 60's and 70's, less was known about the hexagonal carbon nets that are the layers of graphite. We now know these nets can be isolated and studied on their own (the discovery of graphene in 2004 by Andre Geim and co-workers at Manchester), they can be rolled into tubes (discovery of nanotubes by Iijima in 1991) and they can form into balls (discovery of fullerenes by Kroto and coworkers in 1985). Thus, existing theories did not think to account for buckling or folding of the graphite layers, which we have shown to be important in radiation damage.In addition, electron microscopes were not as powerful then as now: we can get pictures of the layers of graphite in atomic detail. We can detect spectroscopic signatures of different structures from Raman and electron spectroscopy and even perform holography of the polycrystalline graphite with nanometre precision. Finally, the progress in computer software and hardware means that we can calculate exactly the structures that will result from neutrons colliding with carbon atoms by solving the equations of motion of the electrons that hold atoms together. The comparison between the length of a carbon-carbon bond, which is about one seventh of a nanometre, and the length of a typical graphite component (about a metre) is unbelievably large: 7,000,000,000! So we must use different theories for different length scales so that we can combine our understanding from measurements and simulation at every scale in between. Thus we use a multiscale approach to calculate the shape, strength and rigidity of the graphite components taking into account what the neutrons do to individual atoms, to the layers they reside in, to the crystallites and then to the component as a whole.The result will give predictive power to the nuclear utilities and to the designers of the next generation of inherently safe and efficient very high temperature reactors.