 |
| Titel |
Experimental evidence for healing during stick-slip at the bases of ice streams |
| VerfasserIn |
Lucas Zoet, Neal R. Iverson |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250127943
|
| Publikation (Nr.) |
 EGU/EGU2016-7875.pdf |
|
|
|
|
|
| Zusammenfassung |
| The Whillians Ice Stream has twice daily stick-slip events of ca. 50 cm with a maximum
inter-event time of ca. 60,000 s. In order for stick-slip phenomena to occur under rate and
state friction, two conditions need to be met: 1) A rate-weakening material at the interface, so
that a nucleated slip perturbance can be propagated and 2) a material capable of healing (i.e.,
becoming stronger) when stationary, so that stress can be recharged during hold periods
between ruptures. Although rate weakening has been experimentally demonstrated for some
basal tills, experimental data relevant to glacier slip that bear on healing have been absent.
Without an understanding of the healing mechanisms active at the beds of ice streams,
models of the mechanics of ice stream stick-slip or ice stream shut-down will be inadequately
informed.
We investigated healing mechanisms with slide-hold-slide experiments, a technique
common in rock mechanics, using two different ring shear apparatuses. In one set of
experiments till alone was sheared, while in another set ice at its melting temperature was slid
over till. These two kinds of experiments allowed for the isolation of mechanisms active at
ice-till interface from those within the till. In all experiments sliding velocity was ca. 345
m/yr, and effective stress was ca. 150 kPa. Once steady-state sliding friction, μss, was
attained, sliding was stopped and the materials were held in stationary contact for a given
duration. When sliding was reinitiated, slip resistance initially rose above the previous μss
value to a peak friction, μpeak, before returning to μss. The difference between μss and
μpeak, Δμ, was then calculated. For each subsequent hold, the duration of stationary contact
was increased logarithmically (100, 1,000, 10,000 and 100,000 s) until the maximum hold
duration was attained. From the relationship between hold time and Δμ, a healing rate was
calculated.
Results from both sets of experiment indicate that Δμ increases with the logarithm of
hold time but that in experiments with the ice healing rate is significantly faster than in
experiments with till alone at hold times greater than ∼1000 s. This result indicates that some
strengthening is the result of ice-till interactions. Mechanisms postulated previously to
explain strengthening at the ice-bed interface are insufficient for characterizing
the amount and form of strengthening observed over realistically brief hold times.
A new model was generated to relate healing rate to changes in the real area of
contact, Ar, between the ice and bed. The model attributes growth of Ar to the
closure of micro-cavities formed in the lee of grains that span the ice-bed interface.
In the model, Ar increases with the square root of the hold time, owing to cavity
shrinkage that begins when sliding stops. This model yields a reasonably good
fit to strengthening observed in the experiments and provides the first empirical
justification for strengthening between slip events at the bases of soft-bedded ice streams. |
|
|
|
|
|
|