lower layer (Fig. 10). This variation in structure
urea-doped ice, but it reputedly has more realistic
over the ice thickness results in nonhomogeneity
fracture-toughness performance, and thus crack-
of the mechanical properties of the ice, so care
ing replication, because it is nearly single-layered
must be taken when reporting results of how tests
(Fig. 11).
were conducted and how measurements were
A problem with doped ices is their dispropor-
made. The thickness of the upper congelation lay-
tionately large brine content for low strength con-
er can be minimized by seeding and growing the
ditions. The criteria for buoyancy similitude may
ice sheet at the lowest temperature that can be
not be met, because the model ice is denser than ice
achieved.
at full scale. Spencer and Timco (1990) describe a
While saline-doped ice appears not to have
method to incorporate minute air bubbles into a
been used for hydraulic modeling, urea-doped
growing ice sheet, thereby controlling its overall
ice has been used with success in many hydraulic
density. They refer to this ice as controlled density
model studies. Deck (1985) used urea-doped ice
(CD) ice. They can reduce ice sheet densities to spe-
cific gravity values of 0.83 to 0.93 by adding mi-
crobubbles at various times during the growth of
the ice sheet. They report an increase in E/σf of 50
Figure 10. Urea-doped ice thin section
Figure 11. Thin section of EG/AD/S ice (on 1-cm grid) show-
showing the thin upper congelation layer
ing its nearly single-layered composition. (Photo courtesy of
G.W. Timco, National Research Council of Canada.)
and columnar lower layer.
to reduce the strength of a model ice sheet in a
study of a breakup ice control structure. He was
behavior and ice-piece size can be made more real-
able to properly scale the ice strength even
istic by adding microbubbles to sections at the top
though he found it necessary to introduce a dis-
and bottom of the ice sheet only. They also note that
tortion of 4 to provide adequate model depths
the opaqueness of CD ice improves the viewing of
and reasonable model ice thicknesses.
cracking and under-ice movement during testing.
EG/AD/S ice is a model ice material that was
Fine-grained ice. Fine-grained model ice was de-
developed at the Hydraulics Laboratory of the
veloped to further improve the strength property
National Research Council of Canada with the
modeling of weakened ice. A disadvantage of the
goal of producing a model ice that would be sin-
urea-doped ice described above is its double-layer
gle-layered and columnar in structure; i.e., the
composition--a strong upper layer and a weaker,
model ice would not include a congelation layer.
thicker columnar layer beneath. This composition
Timco (1986) reviews the requirements for the
resulted in flexure strengths about different axes
new ice material and describes how the EG/AD/
that differed much more than did the flexural
S combination of chemical dopants was selected.
strengths of most full-scale ice sheets. The lower
The three dopants are ethylene glycol, aliphatic
layer also resulted in some residual strength after
detergent, and sugar; hence the name EG/AD/S.
initial breakage. An important advancement pro-
This weakened ice produces flexural strength and
duced by fine-grained model ice is its homogeneity
E/σf ratio values that are very close to those of
of composition and strength.
13
Previous Page
