Content uploaded by Sergio Llana-Funez
Author content
All content in this area was uploaded by Sergio Llana-Funez on Jan 09, 2023
Content may be subject to copyright.
Use of pore volumometry to study the interaction between deformation and
metamorphism in experimental dehydration of serpentinites and gypsum
Uso de volumometría de poro para el estudio de interacciones entre deformación y metamorfismo
en la deshidratación experimental de serpentinitas y yeso
S. Llana-Fúnez1, E. Rutter2, Dan Faulkner3, Kate Brodie2 and John Wheeler3
1. Dpto. Geología, Facultad de Geología, Universidad de Oviedo 33005 Oviedo (Asturias). slf@geol.uniovi.es
2. School of Earth, Atmospheric and Environmental Sciences, University of Manchester. M13 9PL Manchester, Reino Unido
3. Earth and Ocean Sciences, School of Environmental Sciences, University of Liverpool. L69 3GP Liverpool, Reino Unido
Abstract: The effects on mechanical properties of rocks as a consequence of the progress of mineral
reactions are particularly accessible and relatively easy to monitor in laboratory experimentation in the
case of dehydration reactions because of the bulk volume changes involved as well as in the solid
fraction. Hydrated minerals are primary rock-forming phases in fault rocks or in rocks associated with
large structures, both in oceanic and in continental crusts and, for that reason, they exert a major control
on their evolution. Hence the interest in their study.
In this contribution we review the main features of the experimental technique that we use, pore
volumometry, the type of data that can be obtained at moderate pressures and temperatures (200 MPa,
700 ºC ), and outline some of the processes studied to date.
Key words: Rock deformation, experimental petrology, microstructure, porosity, dehydration reactions.
Resumen: Los efectos en las propiedades mecánicas de rocas producidos por el desarrollo y el avance
de las reacciones minerales son particularmente reproducibles y monitorizables en el laboratorio en el
caso de reacciones de deshidratación debido fundamentalmente a los cambios sustanciales de volumen
que se producen en las muestras, de manera global y afectando a la fracción sólida. Los minerales
hidratados son fases minerales mayoritarias en rocas de falla o asociadas a grandes estructuras en la
corteza tanto continental como oceánica y ejercen un control sobre la evolución de las mismas, de ahí el
interés en su estudio experimental.
En esta contribución revisamos los aspectos más interesantes de la técnica experimental, volumometría
de poro, el tipo de datos que se pueden obtener a presiones y temperaturas intermedias (<200 MPa, 700
ºC) y mostramos algunos de los resultados obtenidos hasta la fecha.
Palabras clave: Mecánica de rocas, petrología experimental, microestructura, porosidad, reacciones de
deshidratación.
INTRODUCTION
The implications of variations in the pore fluid
pressure inside porous rocks are known for some time
from the studies of Terzagui (1943). He used the
concept of effective pressure to refer to the difference
between confining pressure, equivalent to burial depth
or the mean stress supported by the solid framework,
from the pore fluid pressure, and related effective
pressure with the mechanical strength of the material.
Originally defined for soils and non-consolidated
geological materials, the implications for low porosity
crystalline rocks were considered some time later
through two seminal papers by Hubbert and Rubey
(1959) and Raleigh and Paterson (1965). The first
paper argued that pore fluid pressure excess,
consequence of the dehydration reaction of gypsum,
contributes to generate the low friction necessary to
allow low angle thrusts the accommodation of large
displacements recorded in foreland fold and thrust
belts. The second paper explored experimentally the
mechanical consequences of the pore fluid pressure
increase produced when fluid expelled from
dehydration reaction of serpentine is not allowed to
leave the sample. More than fifty years later, not all
aspects of dehydration reactions at moderate pressures
are fully understood, particularly those related with the
effects on thermodynamics and the progress of
metamorphic reactions. Here we present a recent
account of the experimental results that can be obtained
using pore volumometry in a triaxial deformation
apparatus in natural rock specimens that undergo
dehydration reactions.
EXPERIMENTAL METHOD
The experimental approach to the study of
dehydration reactions at moderate pressures resides in
the volume changes associated with the reactions.
Figure 1 shows that the reduction of solid volume
consequence of the reaction generates porosity which,
under a substantial pressure interval, is not able to
accommodate the volume of fluid released during the
reaction. If bulk volume is kept constant
experimentally, fluid pressure will build up to the point
that the pore fluid may reach lithostatic or potentially
even higher pressure. This reduces effective pressure
and produces mechanical embrittlement, as shown by
Raleigh and Paterson (1965). If fluid pressure increases
over the tensile strength of the rock it may lead to
hydrofracturing.
FIGURE 1. Sketch summarising volume changes associated with
dehydration reactions. For lizardite dehydration the volume of water
evolved is about 45 % of the initial volume, 25 % is hosted by the
porosity generated during the reaction, the 20 % remaining is the
water expelled if pore fluid pressure is kept constant (approximate
figures for dehydration at 560 ºC and 120 MPa). For gypsum,
volume of water produced is 37 %, 28 % is hosted by the porosity
generated during the reaction and 9 % expelled from the samples if
drained. These figures correspond to experiments under no effective
pressure, when pore fluid pressure equals confining pressure.
Generally, pore fluid pressure in nature will be a fraction of the
confining pressure, in which case the rock (solid+pore space) will
compact, increasing the amount of fluid expelled from the samples
(case on the right in the figure).
FIGURE 2. View of deformation apparatus at the University of
Liverpool. Some of the main components are indicated in the picture.
As a reference for scale the with of the pressure vessel is 10 cm.
In the laboratory, pore fluid pressure can be kept
constant, provided pore fluid connectivity is sufficient.
In that case, mechanical properties will depend mainly
on the fraction of the porosity generated during the
reaction. As the porosity increases the rock becomes
subject to compaction if effective pressure is
sufficiently high and we have shown before that this
applies to experimentally dehydrated serpentinites
(Rutter et al., 2009). Pore fluid access to the samples is
key in experimental studies. Sample pore space can be
reached following two routes, from the top of the
specimen (upstream) and through the bottom
(downstream), either simultaneously (connecting
externally both reservoirs) or individually (isolating
externally both reservoirs). The deformation apparatus
in Figure 2 uses cylindrical specimens that are hosted
inside a long pressure vessel. Long vessels allow to
heat samples externally and to have sealing parts away
from the heaters (additional cooling systems can be
installed to prevent the heat damaging the seals). The
rig shown in Figure 2, located at the University of
Liverpool, works to relative low T (200 ºC) but
illustrates the principles followed in higher T rigs, such
as those used in the dehydration of serpentine (500-700
ºC) at the University of Manchester. The main
advantage of having these two routes of access to the
specimens is to allow the continuous measurement of
permeability through the rock specimen. Although
permeability was possible to measure in earlier
designs, the recent improvement is that the record is
continuous and that it can be done while other
experimental values change, as long as changes are
small (e.g. Mitchell and Faulkner, 2008).
METAMORPHIC REACTIONS
We have studied two dehydration reactions:
gypsum to bassanite, at low temperature, and
serpentine (lizardite) to olivine and talc, at intermediate
temperature (>520 ºC). In both cases, the bulk volume
excess due to expulsion of water is substantial (Fig. 1)
and accesible to continuous monitoring. Gypsum
experiments were done at the University of Liverpool,
and serpentine experiments at the University of
Manchester. At laboratory confined conditions gypsum
dehydrates to bassanite (Fig. 3a) and releases about 9
% of water after full dehydration in absence of
compaction (Olgaard et al. 1995). Lizardite dehydrates
to olivine plus talc and releases up to about 20 % of
water (see Fig. 1 and 3b), very much depending on the
water molar volume (experiments are done in
conditions where water is supercritical steam) (Llana-
Fúnez et al., 2007). In both cases, a set of experiments
were done first in conditions where the pore fluid
pressure equals confining pressure, so that the effective
pressure is close to nil and the porosity is preserved
after the experiments (Fig. 3). In such cases, the state
of stress is hydrostatic and the reaction is fully
described by hydrostatic thermodynamics.
MECHANICAL DATA
There are several types of deformation experiments
that can be run in triaxial deformation apparatus. In the
case of porous rocks, the simplest way to deform a rock
is to apply effective pressure, that eventually may lead
to compaction of the porosity. Other arrangements
require the piston to be displaced upon the specimen.
Figure 4 shows an example of the data obtained when
the samples are loaded at constant velocity of the
moving piston. For short displacements and small
strains, this type of experiments approximate constant
strain rate experiments. In dehydrating materials, the
main two effects that can be observed are the reduction
of the yield point, which defines the upper stress limit
of elastic behaviour in rocks, and the strain hardening
post-yield (the change in slope of differential stress
versus strain with respect to what should be a perfect
plastic material). The steep slope beyond the yield
point in dehydrated and dehydrating materials
illustrates the strain hardening associated with the
compaction of the porosity as the sample deforms. It is
a typical behaviour in high porosity rocks and is also
found in dehydrated serpentinites with sufficient
porosity (e.g. Rutter et al. 2009).
FIGURE 3. (a) Bassanite needles product of dehydration of gypsum
at moderate pore fluid pressure. Top micrograph view was taken
under optical transmitted light with parallel nicols, bottom with
crossed nicols. (b) Backscattered SEM micrograph showing small
and elongated olivine grains (fo) growing in optical continuity. Talc
thin sheets can be seen growing in between the olivine grains.
Starting material has relic fragments of olivine grains that can be
seen in the bottom right.
An alternative approach to relate reaction progress
in rocks under differential stress are creep experiments
in which a constant load is applied while the rocks
undergo metamorphic changes. This configuration is
common in high temperature equipment (where on the
other hand there is no control on fluid pressure),
however few experiments in experimental dehydration
have been published to date.
To extrapolate mechanical behaviour recorded in
laboratory tests to strain rates closer to nature, it
requires to know among other things the strain rate
dependence of deformation processes operating in
dehydrating rocks. Relaxation experiments allow to
explore the mechanical behaviour of rock specimens
over several orders of magnitude of strain rate, as low
as 10-8 or 10-9 s-1. For this type of experiment the
sample is loaded at constant velocity up to the target
stress, at which point the piston is stopped and the rig
and sample are let to relax over time. Strains are small
but are accommodated over long periods of time.
FIGURE 4. Reduction of the yield point and characteristic post-yield
strain hardening behaviour related with the generation and
collapase of porosity during the dehydration of serpentine. Samples
come from serpentinized peridotites from the Herbeira massif in
Cabo Ortegal (NW Spain).
DEFORMATION AND METAMORPHIC
PROCESSES
Experimental conditions during separate
metamorphic and mechanical studies cannot be more
different. Equilibrium during dehydration reactions
require fluid pressure to be equal to lithostatic.
However, under those conditions of hydrostatic
pressure or isotropic stress, the strength of the material
is low, high pore fluid pressure will reduce friction at
grain boundaries. In fact, in nature, only in some
specific settings and situations pore fluid pressure will
equal lithostatic pressure. On the other hand,
dehydration experiments at high effective pressure will
promote the compaction of the porosity that is
generated during the reaction, masking partially the
progress of the reaction when tracked using pore
volumometry.
In the case of drained experiments, when the pore
fluid pressure is kept constant, the strength of the
reaction product is determined by the porosity fraction.
We have shown that the mechanical behaviour of
experimentally dehydrated serpentinites can be
described by critical-state soil mechanics (Rutter et al.,
2009). The implications for fault zones outlined by
serpentinites is that these rocks under these conditions
cannot accumulate stresses sufficient to nucleate
earthquakes, as opposed to previously thought.
However, the fluids release and drained can promote
fracturing in neighbouring stronger rocks (e.g.
gabbros).
The study of dehydration reactions with separate
control on pore fluid pressure and confining pressure
has also allowed us to explore what the contributions of
these two parameters are to the progress of the
dehydration reaction. Using intact specimens of
polycrystalline gypsum from Volterra we have shown
experimentally that pore fluid pressure does have a
major control on the progress of the reaction than
confining pressure has (Llana-Fúnez et al. 2012).
Although this was assumed in earlier works (e.g. Miller
et al., 2003) it had not been demonstrated
experimentally.
Both sets of experiments illustrate the need to
combine both aspects of tectonic processes in rocks
that occur in deep crustal environments, whether on
continental or oceanic crust.
ACKNOWLEDGEMENTS
The research was supported by UK NERC Grants
NER/A/S/2003/00305 and NE/C002938/1. Additional
funding to SLF during the writing of the experimental
results is also acknowledged from Spanish MICINN
grants RYC-2008-02067 and CGL2010-14890. We
thank J.I. Gil Ibarguchi for his review.
REFERENCES
Llana-Fúnez, S., Brodie, K.H., Rutter, E., y Arkwright,
J.C. (2007). Experimental dehydration kinetics of
serpentinite using pore volumometry. Journal of
Metamorphic Geology, 25: 423-438.
Llana-Fúnez, S., Wheeler, J. y Faulkner, D. (2012).
Metamorphic reaction rate controlled by fluid
pressure not confining pressure: implications of
dehydration experiments with gypsum.
Contributions to Mineralogy and Petrology, In
press.
Miller, S., van der Zee, W., Olgaard, D.L., and
Connolly, J.A.D. (2003): A fluid-pressure feedback
model of dehydration reactions: experiments,
modelling, and application to subduction zones.
Tectonophysics, 370: 241-251.
Mitchell, T. and Faulkner, D.R. (2008): Experimental
measurements of permeability evolution during
triaxial compression of initially intact crystalline
rocks and implications for fluid flow in fault zones.
Journal of Geophysical Research, 133: B11412, 16
p.
Olgaard, D.L., Ko, S.C. and Wong, T.F. (1995):
Deformation and pore pressure in dehydrating
gypsum under transiently drained conditions.
Tectonophysics, 245: 237-248.
Rutter, E.H., Llana-Fúnez, S., Brodie, K.H. (2009).
Dehydration and deformation of intact cylinders of
serpentinite. Journal of Structural Geology, 31: 29-
43.
Terzagui, K. 1943. Theoretical Soil Mechanics. John
Wiley, New York, 510 p.
