Sensitivity of Saffman-Taylor fingers to channel-depth perturbations

Andres Franco Gomez, Alice Thompson, Andrew Hazel, Anne Juel

    Research output: Contribution to journalArticlepeer-review

    Abstract

    We examine the sensitivity of Saffman–Taylor fingers to controlled variations in channel depth by investigating the effects of centred, rectangular occlusions in Hele-Shaw channels. For large occlusions, the geometry is known to support symmetric, asymmetric and oscillatory propagation states when air displaces a more viscous fluid from within the channel. A previously developed depth-averaged model is found to be in quantitative agreement with laboratory experiments once the aspect ratio (width/height) of the tube’s cross-section reaches a value of 40. We find that the multiplicity of solutions at finite occlusion heights arises through interactions of the single stable and multiple unstable solutions already present in the absence of the occlusion: the classic Saffman–Taylor viscous fingering problem. The sequence of interactions that occurs with increasing occlusion height is the same for all aspect ratios investigated, but the occlusion height required for each interaction decreases with increasing aspect ratio. Thus, the system becomes more sensitive as the aspect ratio increases in the sense that multiple solutions are provoked for smaller relative depth changes. We estimate that the required depth changes become of the same order as the typical roughnesses of the experimental system ( 1 μm ) for aspect ratios beyond 155, which we conjecture underlies the extreme sensitivity of experiments conducted in such Hele-Shaw channels.
    Original languageEnglish
    Pages (from-to)343-368
    JournalJournal of Fluid Mechanics
    Volume794
    Issue number0
    Early online date4 Apr 2016
    DOIs
    Publication statusPublished - 1 May 2016

    Fingerprint

    Dive into the research topics of 'Sensitivity of Saffman-Taylor fingers to channel-depth perturbations'. Together they form a unique fingerprint.

    Cite this