This paper presents initial results from scaled geotechnical centrifuge mod
elling of cryogenic slope processes. Four experiments are described in whic
h 1/10 scale planar slope models were constructed from a silty soil at grad
ients of 12 degrees, 18 degrees and 24 degrees. Models were frozen on the l
aboratory floor, and thawed in the centrifuge at 10 gravities. Frost heave,
thaw settlement, soil temperature and pore water pressures were recorded.
In each experiment, ten columns of 5 mm long plastic cylinders inserted thr
ough the soil profile allowed surface soil displacements to be determined,
and indicated displacement profiles with depth at the end of each experimen
t. The 12 degrees model was subjected to four cycles of freezing and thawin
g, simulating four annual active-layer freeze-thaw cycles. During each thaw
phase, gelifluction occurred, and average model-scale surface displacement
s were 18.7 mm/cycle (equivalent to 187 mm/cycle at prototype scale). In th
e 18 degrees model, gelifluction rates were higher, with average surface di
splacements of 65.1 mm/cycle at model scale, 651 mm/cycle at prototype scal
e. Two replicate 24 degrees frozen slope models were tested in the centrifu
ge, and thawing was associated with failure by mudflow. Maximum pore water
pressures during thaw were similar in all models. Slope stability analysis
using an infinite planar model indicated that the factor of safety against
failure remained >1 in the 12 degrees model, was close to or slightly less
than 1 in the 18 degrees model, and was <1 in the 24 degrees models. Copyri
ght (C) 2001 John Wiley & Sons, Ltd.