30 / 80
30
Annual Report 2016
SAMCoT
During 2016, PhD candidate Åse Ervik finished analyzing
the data on ice ridge measurements carried out during
the N-ICE2015 expedition. By the end of 2016, Ervik and
co-authors submitted two journal papers for review.
Ervik studied the development of the rubble ma-
cro-porosity and the strength of the consolidated layer.
The key findings were that the rubble macro-porosity
continuously decreased, while the consolidated layer
thickness increased, during the lifetime of a saline ridge.
The strength of the consolidated layer decreased until
the ice reached an isothermal state, after which the
strength remained constant even when the brine volume
increased. Additionally, the decaying ice exhibited ductile
behaviour. The implication of these results to structures
in icy waters, is that decaying ridges could cause severe
loads even if the ice strength is low when compared to
cold ice. The combined effect of both ductile ice and a
PROPERTIES OF DECAYING ARCTIC RIDGES AND MODELLING OF CONSOLIDATED ICE
Numerical modelling of ice crushing. Simulation result of
a large deformation continuous ice crushing experiment
conducted in Svea 2003 (Moslet et al. 2004). A symmetric
boundary condition was used to simulate one half of the
cylinder and ice sheet. The white cylinder was considered
fixed and rigid. The colours represent the Von Mises stress
distribution, where the red colour shows the maximum
stress and the blue is the minimum.
CONTINUUM MODELLING OF ICE RUBBLE
During 2016, Sergey Kulyakhtin studied the behavior of
unconsolidated ice rubble. The role and behavior of ice
rubble is very important in scenarios where ice ridges
interact with both fixed and floating structures.
The consolidated part of an ice ridge is much stronger
than the unconsolidated part, which is called ice rubble.
The ice rubble, though weaker, can be 20 times larger
and can therefore produce a significant load. Due to the
large size of the ice rubble, its total load is difficult to
measure. However, basin-scale measurements can be
used to validate numerical models which can provide
information for larger scale scenarios. Hence, the
importance of numerical models.
When ice rubble interacts with a structure ice fragments
within the ice rubble can break. This complex behavior
is difficult to incorporate into discrete models. Kulyakh-
tin used a continuum model, developed in WP2, to study
the effect of ice breakage in ice rubble. He studied and
modelled the process of ice rubble-structure interaction
comparing his results to measurements obtained in
lab experiments performed at the Hamburg Ship Model
Basin during the RITAS project in 2012.
The most interesting result from the model predictions
was that the amount of ice breakage in ice rubble
had only a minor effect on the resulting load. Hence,
the hardening of ice rubble due to breakage does not
increase the load from floating ice rubble. This could
be not the case if the ice rubble were grounded. In
contrast, the accumulation of ice rubble appeared to be
the decisive factor for the rubble load. The results of the
study are described in a paper submitted to the Journal
of Cold Regions Science and Technology.
In addition, Kulyakhtin continued his study on the
uncertainties involved in approximating ice rubble by
continuum models in collaboration with Arttu Polojärvi
from Aalto University. They quantified the error involved
in using the stress tensor instead of contact forces
between blocks. This error is due to non-uniform force
distribution (force chains) in the ice rubble sample.
Kulyakhtin and Polojärvi also studied the effect of this
error on the angle of internal friction, which is widely
used to characterize ice rubble. When the sample size is
increased from five block lengths to 20 block lengths,
the error in the stress ratio decreases from 31% to
9%. The same increase in the sample size reduces the
error in the internal angle of friction from 12% to 3%.
Further information on the results of this study will be
presented at the POAC17 Conference.
In this plot we can see that the accumulation of ice
rubble in front of the structure in four time instances is as
predicted by the model and measured from the underwater
cameras.
Ice rubble load predicted by the model with different parameter
sets and measured in the ice tank. Here we see that most
of the parameters show only a minor effect on the load. The
strongest effect is seen in Test 6 which corresponds to the
highest ice rubble accumulation.
thick consolidated layer may affect the governing failure
mode and result in severe global forces.
In addition, Ervik worked on modelling the crushing
failure of the consolidated layer. Ervik tested different
numerical formulations and constitutive models for
simulating the process of ice crushing failure against a
vertical fixed structure. The results from these tests will
be presented at the POAC Conference in 2017.