Page 281 - Caldera Volcanism Analysis, Modelling and Response
P. 281

256                                                            J. Martı ´ et al.

















          Figure 10  Sketch of the thermo-mechanical model showing initial and boundary conditions
          (modi¢ed after Guillou-Frottier et al., 2000).

             models (e.g. Gray and Monaghan, 2004) is that one can track the nucleation
             and growth of fractures. Alternatively, the influence of thermal effects on the
             stress field can be explored (Burov and Guillou-Frottier, 1999; Guillou-
             Frottier et al., 2000), since these effects can alter crustal rheology and hence
             influence the formation and subsequent development of fractures. In general,
             non-elastic models dealing with overpressure load condition divide the
             collapse process into two different stages. First, the chamber’s overpressure
             triggers ground uplift, roof bending, fracturing, and magma extrusion, and
             second, erupted materials accumulate. When the sum of chamber excess
             pressure and roof strength can no longer balance the load of the erupted
             products, the chamber roof starts to flex down and subsides. Some models
             cover both stages of the collapse (e.g. Burov and Guillou-Frottier, 1999)
             while others simply focus on the development of fractures during the second
             phase (e.g. Gray and Monaghan, 2004). The calculations provide stress and
             thermal regimes versus time around the magma chamber and predict fault
             location. For example, Burov and Guillou-Frottier (1999) find that during
             uplift, overpressure results in flexural uplift of the roof causing bending and
             eventually failure at the borders and initiation of normal inclined border
             faults. The area affected by posterior subsidence of the roof is thus limited to
             the inward-dipping cover. A later snapping of the roof triggers a piston-like
             subsidence. Consequently, this model predicts two groups of faults: inclined
             primary (initiated during the bending stage) and subvertical secondary
             (initiated during overloading and subsidence). Inclined normal faults may be
             initialised either at the surface during the subsidence phase or at depth during
             the possible uplift phase (in this case, they first appear as inverse faults) and
             propagate upward to the surface. Again, magma chamber geometry is a key
             parameter because the number and location of faults depend on the chamber
             aspect ratio. For large aspect ratio chambers (W3), the flexural stress
             concentrates at the upper corners of the magma chamber resulting in the
             formation of inverse inclined border faults with an inclination controlled by
             the friction angle. More eccentric geometries can, in addition, create internal
             embedded faults.
   276   277   278   279   280   281   282   283   284   285   286