|
Fig. 2 Example of stress-strain curves for three different strain rates, (a):

6.1x 10-5 s-1 , (b): 2.7 x 10-4 s-1 , (c): 2.3 x 10-3 s-1 ; test temperature: -15ºC.

be typical for the intermediate range between the purely
ductile and brittle behaviour. The curve shows that duc-
tile failures, causingmicrostructuraldamage, aregoing on,
but the sample finally fails catastrophically after a certain
amount of deformation. The testdurationis typically about
20 s. Curve (c) shows the result of a fast test (strain rate: 2.3
x 10 -4 s-1 ). The type of failure is brittle; the sample breaks
after very little deformation within fractions of seconds
and exhibits minimal toughness.
The strainto failure decreaseswith increasingstrain rate
in the ductile range and seems to be independent of test-
ing ratein the brittle range(Fig. 3). The ductile-brittle tran-
sition is at about 5 x 10-4 s-1 . Typical values for the failure
strain are 1 to 3 % in the ductile range,and 0.05 to 0.1% in
the brittle range.

Temperature effect

The temperature effect (Fig. 4) indicates substantially in-
creasing stiffness and slightly increasing strength for de-
creasing temperature. The change in critical strain is not
typical, since the analysis of all data suggests that the fail-
urestrain is nottemperaturedependent.Preliminary analy-
sis (N = 46) shows that the stiffness increases about 60%
whenthe temperaturedecreasesfrom -5 to -15 ºC. The shear
strength increases about 20% when the temperature
decreases from -5 to -15 ºC. Although the correlation is
significant (N = 46, R = 0.62, p = 0.015), the increase in
strength is of the same order of magnitude as the scatter in
the strength data due to the nonuniformity of the samples
(S.E. 23%).
DISCUSSION

Generally, the results may be only valid for the snow type
and the test equipment used. However, the typical
behavioral trends for snow, known from other laboratory
test experiments (e.g. Narita, 1980; Fukuzawa and Narita,
1993), were observed: in particular the rate dependence,
the type of mechanical behaviour and the ductile-brittle
transition, over test durations from tenths of a second to

130
|
 |
|
several hours. Stress-strain curves in the brittle range are
similar to the ones in-situ measured by Föhn and
Camponovo (1996).
The tests give some idea of the effect of temperatureon
some important mechanical properties. The preliminary
results show astrongincreasein snow stiffness andaslight,
but significant, increase in snow strength with decreasing
test temperature. The larger stiffness at colder temperatures
suggests smaller deformations, and consequently the
release probability decreases which is consistent with
explanationsof slab avalanche formation reportedin detail
by McClung (1996) and summarized by McClung and
Schweizer (1996).
|
 |
|
Fig. 4 Example of stress-straincurves for three different test tempera-
tures: -5, -10, and -15 ºC. Snow stiffness in these examples is about
200, 300 and 600 MPa, for -5, -10, and -15 ºC, respectively.
|
 |