Capillary Tension in Rotting Ice Layers
RACHEL E. JORDAN AND JEFFREY A. STARK
metically sealed Tempe cell (Fig. 2) and then subjected
INTRODUCTION
to incremental increases in air pressure to "push" off
Ice layers within snow or at the ground interface
the water. To control the ice vein width, and hence the
impede and divert the flow of water, thus leading to
permeability and entrance pressure, we submerged the
ponding and flooding during rain-on-snow events or
ice sample in a refrigerated bath maintained just below
0C. The setup is kept within a coldroom maintained
during springtime thaws. Thawing of the basal ice layer
near +5C. The Tempe cell consists of a brass cylinder
affects terrain and road mobility. We must therefore
question how ice behaves in contact with soil and in
and two plastic end pieces, sealed airtight with pairs of
the context of layered porous media. An upgraded ver-
"O" rings. Water volume removed from the ice is mea-
sion of the SNTHERM snow model (Jordan 1991, Jor-
sured through level changes in a graduated pipette con-
dan et al. 1999) describes movement of water under
nected by plastic tubing to the bottom of the Tempe
thermal and liquid gradients in frozen and unfrozen soil.
cell. A 1-bar semipermeable ceramic plate (Soilmoisture
In this model (SLTHERM), temperature, pressure, ice,
Equipment Corp.) separates the pressurized air cham-
and liquid are coupled through constituent relationships
ber from the open water reservoir. Bubbling pressure
that are universally used for soil (e.g., Black and Tice
and hydraulic conductivity of the plate are specified as
20 to 30 PSI and 3.46 107 cm/sec. Water remained
1989, Black 1990). While water flow through glaciers
is often treated as Darcian saturated flow (e.g., Colbeck
unfrozen within the ceramic stone because of its relatively
1976, Lliboutry 1996), here we consider whether con-
high matric tension. A resistance-wire heater, connected
stituent relationships similar to soil can be applied to
to a temperature regulator, prevented freezeup of water
ice. We ask the fundamental question whether air and
within the submerged tubing.
liquid pathways can co-exist in ice, and if so, what
We used a custom-made refrigerated bath for our
configurations these phases take. In this paper we con-
experiments (Black 1986), consisting of a 20-gallon
centrate on the moisturetension constituent rela-
container that is subjected to continual cooling by a
tionship, which characterizes the airliquid proportion
refrigeration system. Because of its large thermal iner-
in the pore space (or veins) as a function of pressure
tia, this bath is capable of constant temperature condi-
tions ( 0.003C) for time durations on the order of
drop across the air/liquid interface (i.e., capillary ten-
sion). This work was in part prompted by Mader's (1992
weeks. A Bailey temperature controller is used to apply
a, b) theoretical and experimental study of equilibrium
the necessary heat through an immersed heating ele-
vein widths in temperature ice above 0.5C.
ment to bring the bath to the desired temperature.
Although this approach is not energy efficient, it works
well because it is easier to precisely warm the water
EXPERIMENTAL PROCEDURE
than it is to cool it. To promote efficient energy
exchange, the bath is continually mixed. We mounted
We measured the moisture characteristic curve (MC)
the Tempe cell apparatus on a freestanding support to
of ice with a procedure similar to that used for soils.
isolate it from the agitation of the bath.
Figure 1 shows a schematic of the experimental appa-
We made experimental runs on two ice samples. The
ratus, in which an ice sample is placed within a her-
Previous Page
