The site investigation for a proposed 300 GeV proton accelerator in the Chalk adjacent to the village of Mundford, Norfolk involved a ground-stability analysis of a 2.4 km diameter accelerator tunnel and 5 tangential beam tunnels 2 to 4km long. The proposed design of the foundations and shielding would have given rise to high vertical ground pressures, but would require minimal differential subsidence (mm per 100s of metres) to preserve the stability of the proton beam. The prohibitive cost of quantitative surveys over even a small part of the 5 x 9km site was such that the only practical option was to devise a qualitative method of assessing the likely load-deformation properties of the Chalk that was applicable to the whole site. To do this, the present author made a geological map of the site, and determined the geological succession and the chemical and mechanical weathering state of the Chalk by visual examination in 80 x 0.75m diameter shafts (Gallois, Figures 2 to 13 in Ward et al., 1968).
When it became clear that the fracture classification (Grades I to IV) devised as part of the geological assessment appeared to be applicable to the whole of the site, a large-scale tank-loading test and an accompanying series of plate-loading tests were carried out at a site adjacent to the ring. These gave consistent results for the bulk mechanical properties of the rock mass that could be broadly correlated with the fracture classification. Much of the succession at the shallow depths required for the investigation comprised lithologically uniform, moderately hard to hard chalks of the Middle Chalk. Beds of more brittle chalk were present at a few stratigraphically predictable levels. When tested for stiffness and creep these showed significantly higher values of Young’s modulus. The fracture classification was therefore modified to include an additional grade to take account of this hardness. The possibility of a visual inspection system that appeared to predict the bulk engineering properties of chalks at little cost proved attractive to engineers. Not surprisingly, it has been used inappropriately at times for lithologically different chalks and in different structural settings. It remains, with modifications, one of the two most widely used classifications for the description of the mechanical weathering of chalks. The CIRIA classification (Lord et al., 2002) is a modified version that seeks to make it more applicable to all chalks