U. S. Army Corps of Engineers

Figure 4-50. Ultimate strength vs. time to failure for Ottawa sand (20-30 mesh) at various temperatures, comparing test

data with computed values87.

and clays indicate that the long-term strength level can

are noticeably curved, that is, shear strength is not

be as low as 5 to 10 percent of that determined from a

directly proportional to normal stress. Therefore, when

standard test for compressive strength.

testing frozen soil for shear strength, the tests must be

d. The load carrying capacity of frozen soil is

performed using a normal stress near the value to which

the in-situ material will be subjected and large

essentially determined by its shear strength and within

extrapolations of the shear envelopes should be avoided.

certain limits the relationship between shear strength and

It is permissible and conservative to use a secant to the

normal pressure may be written as:

s = c + p tan φ

envelope if a straight line relationship between the shear

(Equation 2)

and normal stress is desired, provided the secant

where

c = cohesion, which is dependent upon

intersects the envelope at the in-situ stress value.

both temperature and time

φ = angle of internal friction

f.

Experimental work indicates that long-term

shear strength of saturated frozen soils can be

p = total normal stress.

determined for practical purposes by the equation:

e. To show graphically how shear strength varies with

time, Mohr's envelopes for frozen sand are shown in

σ

β

=

(by Vialov)

(Equation 3)

figure 4-51. The envelopes are plots of the ultimate

t

1n(t/B)

shear strength vs. normal stress for different rates of

applied strain in controlled strain rate type triaxial tests.

where σt

=

the constant stress level at which

Two characteristics are of interest here. First, the

failure will occur at time t in psi

ultimate shear strength is less for the low strain rates

than for the higher ones; and secondly, the envelopes