GETTING GRANITE DIKES OUT OF THE SOURCE REGION

Whether a dike can propagate far from a magma reservoir depends upon the competition between the rate at which propagation widens the dike and the rate at which freezing constricts the aperture available for magma flow. Various formulations are developed for a viscous fluid at temperature T-m intruding a growing crack in an elastic solid. The initial solid temperature equals T-m at the source and decreases linearly with distance from the source. If T-m is the unique freezing temperature of the fluid, dike growth is initially self-similar and an essentially exact solution is obtained; if T-m is above the solidus temperature, the solution is approximate but is designed to overestimate the distance the dike may propagate. The ability of a dike to survive thermally depends primarily upon a single parameter that is a measure of the ratio of the dike frozen margin thickness to elastic thickness. Perhaps more intuitively, one may define a minimum distance from the essentially solid reservoir wall to the point at which the host rock temperature drops below the solidus, necessary for dikes to propagate far into subsolidus rock. It is concluded that for reasonable material properties and source conditions, most basalt dikes will have little difficulty leaving the source region, but most rhyolite dikes will be halted by freezing soon after the magma encounters rock at temperatures below the magma solidus. While these results can explain why granitic dikes are common near granitic plutons but rare elsewhere, the potentially large variation in magmatic systems makes it premature to rule out the possibility that most granites are transported through the crust in dikes. Nonetheless, these results highlight difficulties with such proposals and suggest that it may also be premature to rule out the possibility that most granite plutons ascend as more equidimensional bodies.