Measurement of Properties - Geotechnical Engineering Investigation - page 237

Curve shape significance: A truly undisturbed specimen will yield a curve with a rela-
tively flat initial portion to the range of prestress, then a distinct change in slope near the
maximum p c , and thereafter a steep drop under loads causing virgin compression.
Disturbed samples or samples with a large granular component yield a curve with a grad-
ual change in slope and a poorly defined p c as shown in Figure 3.78.
Methods to determine p c : The Casagrande construction (Casagrande, 1936) is given in
Figure 3.79 and the Burmister construction (Burmister, 1951b) in Figure 3.78. In the
Burmister construction, a triangle from the unload-reload cycle is moved upward along
the e -log p curve until the best fit is found, where the pressure is taken to be p c . The
method compensates for slight disturbances during sampling and laboratory handling,
which cause the initial curve to be steeper than in situ .
Settlement analysis curve correction: Unload-reload cycles, usually made soon after p c has
been identified, provide data for the correction of the settlement curve to reduce conser-
vatism caused by the slight sample expansion occurring during sampling and handling.
The recompression portion of the unload-reload curve is extended back from the initial
curve at p c to form a new curve with a flatter initial slope.

Stress-Strain Relationships
When a series of increasing loads is applied to a clay soil, the amount of compression occur-
ring up to the magnitude of the maximum past pressure or preconsolidation pressure p c is rel-
atively small to negligible. Stresses exceeding p c enter the “virgin” portion of a stress-strain
curve, and substantially greater compression occurs. It may be expressed by the compression
index C c , the slope of the virgin compression or void ratio vs. log 10 time curve expressed as

C c ( e 1 e 2 )/ log( p 1 / p 2 ) (3.63)

C c is useful for correlating data for normally consolidated clays of low to moderate sensi-
tivity, or for computing settlements when data from a number of consolidation tests are
available. It can be estimated from the expression (Terzaghi and Peck, 1967)

C c ≈ 0.009(LL 10%) (3.64)

where LL is the liquid limit.

P c

/2

C c

FIGURE 3.79
Casagrande construction for determining p C . (From
Casagrande, A., Proceedings of the 1st Conference on Soil
Mechanics and Foundation Engineering, Cambridge, MA,
Vol. 3, 1936, p. 60. With permission.)

0.1 1 10 100

Pressure, tsf