gypsum to 2.90 g/cm3 for anhydrite. The total porosity of the
different varieties of gypsum rocks varies from medium to low
porosity while the anhydrites present low to very low porosity. It
is estimated that the mesopores were formed by dissolution while
the micropores may be due to microcracks. With respect to conventional
geomechanical properties, the gypsum is less resistant
and more deformable than anhydrite. The uniaxial compressive
strength of gypsum varies between 4.8 and 44.7 MPa (very low
to moderate strength). Furthermore anhydrite rocks can be considered
medium to high strength (UCS between 62 and 102 MPa).
Table 17 summarizes the geomechanical properties of these rocks.
Fig. 6 shows the relationship between the uniaxial compression
strength and deformability modulus of gypsum and anhydrite
rocks using the classical diagram of Ratio Modulus. It can be seen
that there is an interesting coincidence between the results of
the studied rocks with data published in the literature for similar
rocks. Both sulphatic rock types may be classified as high to medium
modulus but gypsum falls in the field of low to very low
strength while anhydrites are medium to high strength.
Fig. 7 shows the inverse relationship between the uniaxial compression
strength and the total porosity. However, this relationship
is not very clear between the tensile strength to the total porosity
(Fig. 8).
Fig. 9 shows the experimental data plotted in a diagram of
principal stress and in Table 18 the strength parameters of various
criteria including the strength criterion Mohr–Coulomb, Drucker–
Prager, Kim & Lade and Hoek & Brown.
Swelling and creep under uniaxial compression tests were also
carried to study long term geomechanical properties of gypsum
and anhydritic rocks [13]. The results showed that the time-dependent
behavior of rocks of gypsum and anhydrite is related to two
physicochemical processes which can occur in nature coupled:
(a) chemical transformation of gypsum and anhydrite and (b)
creep. They together or separately can justify the order of magnitude
of the slow deformations observed in some underground in
Europe.
In Swelling tests the performed anhydritic rocks present
increased volume mainly by the precipitation of gypsum crystals
on the surface of the sample and not by the mineralogical
conversion anhydrite-gypsum. This would be an isovolumetric
transformation and not isomolar as most authors sustains. Expansion
rates obtained in laboratory vary from 0.6% to 1.1% per year.
Creep tests under uniaxial constant show clearly sections of
primary or transient creep and secondary creep or stationary.
The rheological model that best represents the creep process is
the Burger model. All rocks had a boundary between the primary
and secondary creep that flow approximately coincides with the
stress to critical energy release (damage start). This suggests that
the mechanisms of creep would be related to the processes of
initiation and propagation of microcracks. Creep tests show that
the gypsum, as expected, has higher viscoelastic and irreversible
deformation than anhydrites. Strain rates for both types of rock
in the secondary creep vary from 1 105/day to 1 108/day
[13].
Fig. 10 shows the relationship between the uniaxial compression
strength and deformability modulus of all samples presented
by Vardé and Vendramini and Fabra [1,6]. It is observed that most
rocks considered low to very low strength have a low to medium
ratio modulus and a considerable amount of samples have strength
less than 7 MPa. No clear trend is observed by type of rock and its
geological age. Probably the physical and chemical processes
(decompression and weathering) suffered by the rocks during their
exhumation may have altered its mechanical properties.
Another factor that could affect the mechanical properties at
laboratory scale is the degree of saturation. Fig. 11 shows that