Page 57 - Fundamentals of Physical Volcanology
P. 57

9780632054435_4_003.qxd  12/10/2007  12:18PM  Page 34





                    34  CHAPTER 3



                    The above discussion tacitly assumed that   density. However, it is now understood to be an
                  because the magma in a dike is buoyant the frac-  expression of the combination of a number of rock
                  ture will grow upward. In fact the orientation of   properties, including the tensile strength of the
                  the fracture will be strongly controlled by the state of  rock (i.e., its ability to resist its mineral grains being
                  stress in the rocks that are fracturing. As discussed  pulled apart from one another), the typical size of
                  earlier, the hot rocks in the mantle behave as fluids  the mineral grains, and the difference between the
                  on long time scales. A fluid cannot sustain differ-  pressure in the molten rock waiting to enter the
                  ential stresses – the fluid flows to remove the dif-  fracture when it opens and the compressive stress
                  ferences in stress, and so the force acting within   trying to force the fracture shut. The compressive
                  a body of fluid is the same in all orientations.  stress is partly due to the weight of the overlying
                  However, the cooler rocks in the crust can easily  rock in the place where the fracture is about to
                  resist deformation on short time scales, and so can  form and partly due to any other tectonic stresses
                  support nonzero stress differences. The basic stress  that are present, such as those linked to plate tec-
                  acting on any body of rock is the vertical compres-  tonics. The surface tension between the molten
                  sive force representing the weight of the overlying  rock and the mineral grains is also important. We
                  rocks. The horizontal forces acting may be sig-  saw in section 2.4.2 that surface tension prevents
                  nificantly different from this, since they include   liquid rock spreading in an infinitely thin film over
                  the consequences of horizontal movements within   the surfaces of mineral grains so that, when melting
                  the crust due to, for example, plate tectonics. The  first begins, a finite amount of melt is needed before
                  result is that horizontal stresses in the deep crust  an interconnected melt network can exist. In the
                  are generally found to be about 30% less than the  same way, melt cannot enter an extremely narrow
                  vertical stress. When a fracture opens, the amount  crack at all, and so the extreme tip of an opening
                  of work that it has to do in deforming the surround-  fracture will not contain liquid rock.
                  ing rocks will be minimized if it opens in a plane  One might imagine this implying that there must
                  that is oriented at right angles to the least compres-  be a vacuum in the fracture tip, but in practice all
                  sive stress, and since this is very likely to be hori-  molten rocks contain dissolved volatiles. At low
                  zontal, the plane of the resulting dike will almost  pressures, some of these volatiles will be released
                  certainly be vertical. The situation can be quite dif-  from the liquid rock as gases and can enter the frac-
                  ferent at very shallow depths in the crust, where  ture tip, providing a pressure that is greater than
                  the presence of local layers of rock with low den-  zero, but less than the compressive effect of the

                  sities can cause the least compressive stress to be   overlying rocks. The pressure difference, strictly
                  vertical. In that case, a vertically rising fracture may  speaking a stress difference, between the inside
                  cease to extend upward and may instead spread  and outside acts to try to push the tip of the crack
                  sideways, especially if it finds a weak boundary  shut while at the same time the pressure of the
                  between two rock layers. The approximately hori-  underlying molten rock trying to get into the frac-
                  zontal layer of magma that eventually fills such a  ture is exerting a leverage acting to force it open.
                  fracture is called a sill.                  It is the competition between these two effects
                    Much of the study of the propagation of dikes   that determines the value of the apparent fracture
                  is concerned with two critical issues: what condi-  toughness of the host rocks. However, because the
                  tions allow brittle fractures to start forming, and  apparent fracture toughness depends on the amount
                  what conditions cause the resulting dikes to cease  of volatiles exsolved, which in turn depends on
                  propagating. The rock property controlling these  both the amount of volatiles originally dissolved in
                  processes is called the fracture toughness, and it is  the magma in its source zone and the exact condi-
                  a measure of the maximum intensity of the stress  tions within the crack tip, it is very hard to antici-
                  that the mineral grains at the dike tip can withstand  pate the exact value of apparent fracture toughness
                  before failing. For many years scientists treated   to be expected in a given situation. The units
                  fracture toughness as a basic rock property such as  in which fracture toughness is measured are stress
   52   53   54   55   56   57   58   59   60   61   62