An exploration of thermal counterflow in He II using particle tracking velocimetry

TL;DR

This study uses particle tracking velocimetry to analyze thermal counterflow in superfluid helium, revealing distinct particle velocity regimes and insights into the normal fluid and vortex dynamics.

Contribution

Introduces a novel scheme for separating particle motions influenced by normal fluid and vortex tangle in He II, clarifying previous experimental discrepancies.

Findings

Two distinct peaks in particle velocity PDF at low heat flux

Single peak near half the normal fluid velocity at high heat flux

No size difference between particles in different velocity regimes

Abstract

Visualization of thermal counterflow in He II using PIV (particle image velocimetry) and PTV (particle tracking velocimetry) is difficult because tracer particle motion can be influenced by both the normal fluid and superfluid components of He II as well as the quantized vortex tangle. For instance, an early PTV experiment observed particles moving at the normal fluid velocity $v_n$, while a PIV experiment observed particles moving at $v_n/2$, though the range of heat flux applied in these experiments differed by an order of magnitude. To resolve this apparent discrepancy and explore statistics of particle motion in thermal counterflow, we have applied PTV to a wide range of heat flux at several fluid temperatures. We introduce a scheme for analyzing the velocity of particles presumably moving with the normal fluid separately from those presumably influenced by the vortex tangle. Our results show two distinct peaks in the streamwise particle velocity PDF (probability density function) for lower heat flux, one centered at the normal fluid velocity $v_n$ ("G2") and one near $v_n/2$ ("G1"). For higher heat flux there is a single peak centered near $v_n/2$ ("G3"). Using our separation scheme we show there is no size difference between particles contributing to G1 and G2. We also show that non-classical features of the transverse velocity PDF arise entirely from G1, while the corresponding PDF for G2 exhibits classical Gaussian form. G2 transverse velocity fluctuation, backed up by second sound attenuation in decaying counterflow, suggests large scale turbulence in the normal fluid is absent from the two peak region. We offer a brief discussion of physical mechanisms that may be responsible for our observations, revealing that G1 velocity fluctuations may be linked to fluctuations of vortex line velocity, and suggest numerical simulations that may reveal underlying physics in detail.