Effects of pre-existing fractures on dike propagation: new insights from field data and numerical modelling

To better understand the conditions and mechanisms under which dikes interact with pre-existing vertical fractures, we analyzed the fracture swarm and volcanic vents developed during the 1947 eruption (North-East Rift, Etna), and reconstructed the pre-1947 fracture field by historical aerial photos. Then we developed Finite Element Method numerical models, varying the number, spacing, width and vertical and lateral distance of pre-existing fractures with respect to a dike, as well as the dike overpressure. Although in general the 1947 dike followed the path of a previous dike, at a more detailed scale the 1947 vents are shifted up to 11 m laterally with respect to the nearest pre-existing fracture. In areas where few fractures were already present, a significant number of new fractures developed during the 1947 event, whereas if several fractures pre-existed, a few new ones formed. In the case of a pre-existent single fracture, numerical models suggest two scenarios: nucleation of a new fracture bending toward the pre-existing one, possibly followed by dike deflection, or the vertical propagation of the dike. The latter is facilitated by close pre-existing fractures, because dike/fracture interaction enhances tensile stress above the dike tip. This stress increase is sensitive to horizontal fracture spacing, dike/fracture depth difference, and dike overpressure. Stress concentration towards the dike tip and fracture base means a lower probability of new fractures formation at the surface if pre-existing fractures are already widespread. A dike located nearby, or in the middle, of already existing fractures, receives a stress “booster” enhancing its probability of further propagation.