On the rotation of melting ice disks

Abstract

This investigation was inspired by the work of Dorbolo et al. (Phys Rev E 93(3):15, 2016), which was the first to study, at a laboratory scale, the phenomenon observed in the natural world of floating disks of ice rotating. They conclude that, in controlled conditions, ice disks are able to induce their own rotation. Whilst their work successfully exposes multiple aspects of the kinematics of such disks, and the buoyancy-driven flow generated beneath them as they melt, the mechanism by which rotation is triggered remains unsubstantiated. We therefore return in this work to the study of floating ice disks, focusing specifically on the importance of experimental technique in obtaining reliable measurements of disk behaviour. Our investigation reveals that the motion of ice disks placed on a nominally quiescent body of fresh water is unpredictable, with some disks remaining motionless, and others rotating clockwise or anticlockwise. For those in motion, the average rate of rotation observed was less than half of that recorded by Dorbolo et al., a discrepancy possibly explained by residual background motions in the nominally quiescent surrounding body of water. However, given our observation that non-melting disks of high-density polyethylene (HDPE) cooled to the same temperature as the ice (comparable HDPE and ice disks differing by only sim 4% in mass) consistently rotated at rates less than those made of ice, it is hypothesised that, within the confines of the freshwater environment, the motion of the turbulent meltwater plume that forms beneath an ice disk amplifies the effect of residual background motions. From our observations, it is concluded that residual motions are an underlying physical trigger for disk rotation.