Engineering Rock Mechanics

92 lntact rock

We arrive at the conclusion that if the testing machine stiffness is greater
than the absolute value of the slope at any point on the descending portion
of the stress-strain curve for the rock, the system will be continuously stable
and it will be possible to obtain the complete stress-strain curve. Note also
that, although we have expressed this argument in terms of a uniaxial
compression test, it also applies to tensile tests, shear tests and any other
configuration, such as the stability of a three-point loaded beam test.
The most logical method, therefore, of obtaining the complete
stress-strain curve (from the AEIL equivalent cylinder analogy) is to build
a machine which is large, squat and utilizes only high-modulus materials
(for example, mercury as the hydraulic fluid). One can go even further and
have no moving parts: a solid frame is heated (i.e. expansion takes place),
the specimen is inserted and the frame cooled (contraction results). The
curves in Fig. 6.8 were obtained by this method.
A further point to note, with respect to the curves in Fig. 6.8, is that they
do not all monotonically increase in strain. Invoking the argument used
previously for soft and stiff testing machines, we could not, therefore,
obtain the curve for Charcoal Grey Granite I1 even in an infinitely stiff, i.e.
rigid, testing machine. The stiffness of such a machine would be
represented by a vertical line in Fig. 6.8. To obtain these curves, it was
necessary to further modify the machine using a counter-acting hydraulic
jack in the post-peak region. Wawersik and Fairhurst classified complete
stress-strain curves into two types: Class I curves monotonically increase
in strain; Class II curves do not (Wawersik and Fairhurst, 1970).
Following the pioneering work by Cook, Bieniawski, Fairhurst and
Wawersik in the late 1960s, it was realized that such stiff testing machines
are inherently cumbersome and functionally inflexible. This led to the
introduction of servo-controlled testing machines for obtaining the
complete stressstrain curve for rock. It is important to note that the means
by which a servo-controlled testing machine is able to follow the post-peak
curve is diflerent in principle and implementation from the stiff testing machines.

• Charcoal gray granite I
  --- Indiana limestone
  ...... Tennessee marble I
  -.- Charcoal gray granite I1
  -.. Basalt

] Class I

] Class 11

.......... Solenhofen limestone

.'I

/.

Class I

Strain

monotonically

ob- Infinite increases

stiffness

E

(^0) Class I1
Strain does not
monotonically
increase
Strain, (70) E
Figure 6.8 Examples of complete stress-strain curves for different rocks (from
Wawersik and Fairhurst, 1970).