Transient dike propagation and arrest near the level of neutral buoyancy

We present a finite difference/perturbation method to investigate transient dike propagation from a magma chamber 1–2 km below the level of neutral buoyancy in oceanic crust and explore the role of density gradation in dike propagation arrest. The dike is modeled as a magma-filled blade crack and the host rock is assumed to be a linear elastic medium with graded mass density. An integral equation approach is used to obtain the dike propagation velocity and stress intensity factor at the propagating dike tip. The finite difference/perturbation method is applicable generally to mafic dikes with ηV less than 50 Pa-m, where η is the magma viscosity and V the dike propagation velocity. Numerical results show that dike propagation velocity initially increases with dike length, reaches a peak value when the dike tip reaches a position well above the level of neutral buoyancy and then rapidly decreases to zero after the dike propagates further into the region above the level of neutral buoyancy, which indicates that density gradation is an effective mechanism for dike arrest. The effects of host-rock fracture toughness and magma chamber depth relative to the level of neutral buoyancy are also discussed.