Page 314 - Caldera Volcanism Analysis, Modelling and Response
P. 314
Structural Development of Calderas 289
Table 1 Parameters scaled to simulate the brittle crust and the ductile magma chamber
during caldera-collapse (modified after Merle and Vendeville, 1995).
Parameter Model/nature ratio
Brittle crust
Length L
Density r
Gravity g
Stress s ¼ r g L
Cohesion c ¼ s
Magma
Viscosity m
Strain rate e ¼ s/m
Time t ¼ 1/e
Velocity v ¼ e L
5 3 4
their viscosity m, ranging from 10 (air) to 10 (water) and 10 (silicone) Pa s. The
viscosity of magma can vary over 12 orders of magnitude, depending on its
composition and temperature (Talbot, 1999). Ideally, using a certain analogue
with a given viscosity, it is possible to experimentally simulate a wide range of
magma viscosities, by controlling the related strain rates e (e ¼ s /m ), times t
(t ¼ 1/e ) and velocities of deformation v (v ¼ e L )(Table 1; Merle and
Vendeville, 1995). In practice, all the used magma analogues have some advantage
and limitation. While silicone may be used to simulate very viscous magmas
10
(mB10 Pa s) under ordinary laboratory times (hours), the lower viscosity of air
and, mostly, water remains more suitable to simulate basaltic magmas. However,
unlike silicone, their common confinement within an elastic balloon limits their
flow, mobility and attitude to intrude.
3. Experimental Studies on Calderas
3.1. General features
This section considers the experiments on calderas where collapse is achieved
simulating an underpressure within the analogue magma chamber (Komuro, 1987;
Marti et al., 1994; Acocella et al., 2000, 2001; Roche et al., 2000; Walter and
Troll, 2001; Troll et al., 2002; Kennedy et al., 2004; Lavalle `e et al., 2004;
Belousov et al., 2005; Geyer et al., 2006), including the influence of a regional stress
field (Cailleau et al., 2003; Acocella et al., 2004; Holohan et al., 2005). However, as
part of these 13 experimental sets share common research groups, materials,
apparati and stress conditions, this review will consider only the following eight key
papers: Komuro (1987), Marti et al. (1994), Acocella et al. (2000, 2004), Roche
et al. (2000), Walter and Troll (2001), Kennedy et al. (2004) and Holohan et al.
(2005) (Table 2).