Reconnections of quantized vortex rings in superfluid $^4$He at very low temperatures
Extracted source text ยท arXiv 1308.6171
\documentclass[reprint,aps,prl,showpacs]{revtex4-1}
\usepackage{graphicx,epsfig}
\usepackage{epstopdf}% allows eps figures in pdflatex
\usepackage{bm}% allows bold math
%\usepackage{refcheck}
\newcommand{\4}{$^4$He}
\newcommand{\3}{$^3$He}
%\newcommand{\dif}{{\rm d}}
\newcommand{\delR}{\delta \! R}
\newcommand{\delv}{\delta \! v}
\newcommand{\dx}{\, \mathrm{d}x}
\newcommand{\dif}{\, \mathrm{d}}
\begin{document}
\title{Reconnections of quantized vortex rings in superfluid $^4$He at very low temperatures}
\author{P. M. Walmsley$^a$}
\author{P. A. Tompsett$^a$}
\author{D. E. Zmeev$^{a,b}$}
%\altaffiliation{Present address: Department of Physics, Lancaster University, Lancaster, LA1 4YB, UK}
\author{A. I. Golov$^a$}
\affiliation{$^a$ School of Physics and Astronomy, The University of Manchester, Manchester M13 9PL, UK}
\affiliation{$^b$ Department of Physics, Lancaster University, Lancaster, LA1 4YB, UK}
\date{\today}
\begin{abstract}
Collisions in a beam of unidirectional quantized vortex rings of nearly identical radii $R$ in superfluid $^4$He in the limit of zero temperature (0.05\,K) were studied using time-of-flight spectroscopy. Reconnections between two primary rings result in secondary vortex loops of both smaller and larger radii. Discrete steps in the distribution of flight times, due to the limits on the earliest possible arrival times of secondary loops created after either one or two consecutive reconnections, are observed. %Quantization of the shortest arrival times of secondary charged vortex rings as function of the discrete number of ring-ring reconnections is observed.
The density of primary rings was found to be capped at the value $500{\rm \,cm}^{-2} R^{-1}$ independent of the injected density. This is due to collisions between rings causing piling-up of many other vortex rings.
%initiate large-scale fluctuations in the mean velocity, thus facilitating the transfer of energy from quantum to classical length scales (inverse cascade).
Both observations are in quantitative agreement with our theory.
\end{abstract}
\pacs{67.25.dk, 47.32.cf, 47.27.Cn}% PACS, the Physics and Astronomy
% Classification Scheme.
%\keywords{Suggested keywords}%Use showkeys class option if keyword
%display desired
\maketitle
%\section{Introduction -- reconnections}
Turbulence appears in various systems: fluids, plasmas, interstellar matter -- with common properties such as the existence of long-lived regions of concentrated vorticity, whose reconnections facilitate the evolution of the flow field and redistribution of the kinetic energy between length scales.
%In turbulent flow, regions of concentrated vorticity
%, such as hairpin vortices surrounding wind gusts, are long-lived.
%Their reconnections provide an important mechanism of redistributing the kinetic energy between length scales.
A paradigm of an isolated vortical structure is a vortex ring \cite{ShariffLeonard1992, Donnelly}, and their pair interactions are a testbed of the physics of vortex reconnections.
% In classical hydrodynamics, vortex rings can be thought of as ``elementary excitations'' in turbulence, and understanding their pair interactions might pave a way to understanding more complex processes \cite{ShariffLeonard1992}.
Numerical simulations of collisions of two vortex rings predict various outcomes: either a single ring or several rings, depending on the initial conditions \cite{KoplikLevine96,Kivotides2003}. There were experimental attempts to visualize these processes in classical fluids \cite{Kambe1971, Fohl1975, Oshima1977}; however, they are often hard to interpret because of the inevitable decay, core instabilities and poor characterization of vortex rings in viscous fluids.
Quantized vortex rings in superfluids have an advantage because they are slender and stable, and can hence be well-characterized quantitatively \cite{RR1964}. Recently, there were many theoretical investigations into reconnections of quantized vortex lines \cite{Kerr87, KoplikLevine93, KoplikLevine96, Nore97, Kivotides2001, Leadbeater2001, Leadbeater2003, Chatelain2003, Kivotides2003, Mitani2006, Nemirovskii2006, Barenghi2006, Alamari2008, Kerr2011, Kursa2011, Hanninen2013, Caplan2014}. In particular, for acute angles between two antiparallel reconnecting vortex lines the generation of a cascade of small vortex rings was predicted \cite{Svistunov1995,Kerr2011,Kursa2011}. The outcome is reminiscent of that of the Crow instability \cite{Crow} of antiparallel vortices observed in air. Experimentally, reconnections of vortex lines in superfluid $^4$He have been visualized \cite{Beweley2008,Fonda2014} but only at high temperatures when vortex motion is damped.
Reconnections of vortex loops comprising a vortex tangle, i.\,e. {\it Quantum Turbulence} (QT) \cite{Schwarz1988, VinenNiemela2002}, especially those leading to the emission of vortex rings \cite{Yano2013}, are an important mechanism of redistributing energy towards smaller length scales in QT \cite{Svistunov1995, Tsubota2000, BarenghiRingEmission2002, NazarenkoReconnection, Nemirovskii2008, Nemirovskii2010, KS, Yamamoto2012, Kondaurova2012, Baggaley2013, Salman2013}.
%Reconnections within a tangle with an average inter-vortex distance $\ell$ is expected to produce vortex rings of size $\sim \ell$ as well as cascade of much smaller ones.
% While in classical fluids the lifetime of even an isolated ring is limited due to both viscous damping and instabilities associated with their relatively thick cores \cite{ShariffLeonard1992}, vortex rings in superfuid helium are stable -- thanks to the truly conserved circulation and the absence of dissipation at low temperatures, as well as ultimate slenderness of their cores. This makes superfluids the system of choice for experimental studies of vortex rings \cite{RR1964}.
In superfluid $^3$He-B \cite{Bradley2005} and $^4$He \cite{PRLWalmsley2008}, collisions and subsequent reconnections, in a dense beam of vortex rings generate QT.
%Vortex tangles (i.\,e. {\it quantum turbulence}) can be created by collisions within a beam of vortex rings \cite{Bradley2005,WalmsleyPRL2008}.
Longer and more intensive beams of rings result in tangles that show large-scale velocity fluctuations \cite{TsepelinFluctuations, Roche} and the late-time decay
\cite{Bradley2005,PRLWalmsley2008,WalmsleyLongInjection,WalmsleyPNAS2014}, both characteristic of classical turbulence. This implies the existence of the {\it inverse cascade} of energy from the small length scales (of order ring radii) into which the initial energy is injected -- up to the size of the resulting tangle. It was speculated \cite{Baggaley2013} that the inverse cascade might be maintained by the merger of pairs of rings into larger loops. Yet, no direct quantitative observations of ring-ring reconnections have been reported so far.
%However, the details of any of these processes have never been investigated quantitatively.
%Naive models of tangling of several vortex rings together usually assume their coagulation and do not consider shedding off of smaller vortex rings in the process.
This Letter reports the first quantitative observations of either one or two consecutive reconnections, and the discovery of the ensuing universal state of depleted density -- within a beam of unidirectional quantized vortex rings all of similar radii, with their number density $n$ and radius $R$ under our control. The resulting mechanism of seeding the large-scale velocity fluctuations out of a seemingly random beam of vortex rings is suggested.
In our experiments, to create vortex rings of a required size and to detect their arrival, each was tagged by an excess electron trapped on the vortex core. Applying an electric field along the $x$-axis allowed small seed charged vortex rings (CVRs), injected at $x=0$, to grow to the desired radius $R(x)$, and also to trace the location of the reconnection process that resulted in small secondary charged vortex rings.
The radius of a quantized vortex ring is directly related to its self-induced velocity, $v\sim \kappa/R$ -- which determines its arrival time at the collector at $x=d$. The numbers of primary and secondary vortex rings as a function of their radii could be extracted from the time-dependence of the collector current $I_{\rm c}(t)$ through time of flight spectroscopy.
%we send, from the source to collector, a short beam of unidirectional quantized vortex rings -- each of nearly the same radius and tagged by one electron trapped on its core.
% When the reduced density, $nR^3$, becomes substantial, collisions and hence reconnections occur, resulting in secondary vortex rings of various radii.
The radius of primary rings at the collector, $R(d)$, was varied within 1--6\,$\mu$m, with number density $n(d)$ between $10^{4}$ and $10^{7}$\,cm$^{-3}$; while the mean radius of the seed CVRs was estimated as $\bar{R_0} \leq 0.5$\,$\mu$m.
%To create a beam of vortex rings of similar radii and orientations we would start with negligibly small singly-charged vortex rings and increase their radius to the desired one by applying electric field.
The energy of a CVR, subject to a potential $\phi(x)$, is
${\cal E}(x) = {\cal E}_0 + e \phi(x)$ in the absence of dissipation at $T<0.5$\,K.
The velocity and energy depend on $R$ \cite{Donnelly},
\begin{equation}
v = \frac{\kappa}{4\pi R}\left(\Lambda(R) - \frac{1}{2}\right), {\rm \,\,}
{\cal E} = \frac{\kappa^2 \rho R}{2}\left(\Lambda(R) - \frac{3}{2}\right),
\label{vE}
\end{equation}
where $\kappa = h/m_4$ is the circulation quantum, $\rho$ is the density of superfluid, $\Lambda = \ln\frac{8R}{a_0}$ and $a_0 = 1.3$\,\AA \cite{GlabersonDonnelly1986}.
%The time required to travel from $x=x_1$ to $x=x_2$ is
%The contribution to the collector current at $x=d$, $I(\tau)=en(\pi r^2)v(d)$, will then arrive after the time delay
% \begin{equation}
%\tau = \int_{x_1}^{x_2}v^{-1}\dif x .
%\label{tau12}
%\end{equation}
% For the given $\phi(x)$ and initial energy ${\cal E}_0$ of the ring at $x=x_1$, one can calculate the radius $R(x)$, velocity $v(x)$ and time of flight of a charged vortex ring (CVR) between any $x=x_1$ and $x=x_2$,
A deeper insight can be gained within the approximation for constant $\Lambda \approx 13$ and uniform field $\phi(x) = \frac{U}{d}x$. The radius of a CVR then grows linearly with $x$, \begin{equation}
R(x) \approx R_0 + \frac{2eU}{\rho\kappa^2(\Lambda-3/2)}\frac{x}{d},
\label{R_appr}
\end{equation}
and the time for a CVR to travel from $x=0$ to $x=d$,
\begin{equation}
\tau_1 \approx \frac{4\pi d}{\kappa (\Lambda -1/2)}R_0 + \frac{4\pi e d}{\rho\kappa^3 (\Lambda - 1)^2}U,
\label{tau1}
\end{equation}
increases with energy $eU$ because CVRs slow down as they expand (see Eq.\,\ref{vE}). In what follows, unless specified, we will be using the approximation $R_0 = 0$.
With increasing density of CVRs, collisions become more frequent. These collisions are caused by small fluctuations in the direction and magnitude of the rings' velocities -- mainly due to the variations in initial radii $\delta R_0$ and direction of the seed CVRs when injected at $x=0$. Along with reconnections upon a direct collision, hydrodynamic dipole-dipole interactions between neighboring CVRs (that grows in strength with increasing $n$ and $R$ -- and are hence the strongest near the collector at $x \rightarrow d$) affect the CVR's velocities. The Coulomb repulsion between neighboring CVRs of $R>$\,1\,$\mu$m is much weaker than their hydrodynamic interaction.
% The reconnection of unidirectional rings of similar radii are expected to result in their merger into a larger vortex loop \cite{Baggaley2013}, perhaps accompanied by smaller loops \cite{CommentCollision}.
A reconnection of two CVRs results in secondary vortex loops, which are generally non-circular.
% Yet, provided the smaller-scale perturbations (Kelvin waves) are not excessively large, they can be treated as smoothed circular rings of some effective radius \cite{Maggioni2013}, while the energy of Kelvin waves considered as effectively dissipated \cite{CommentCrabRings}. With this dissipation, the conservation of impulse (which for each CVR is proportional to the square of its radius) requires that one secondary loop has to be larger than the primary ones, while the conservation of angular momentum (proportional to the radius difference, $\sim \delta R_0 \ll R$) requires that at least one loop that is much smaller than the primary ones is created along.
%; thus facilitating creation of strongly unequally-sized secondary loops.
The two trapped electrons are now carried by either one or by two (if any) of the secondary rings.
%There are obviously loops of sizes that are both larger and smaller than the primary loops -- as can be seen by the ensuing broadernning of the collector current peak reflecting the arrivals of charged vortex loops.
% To predict the distribution of the times of their arrivals at collector is a formidable task, because the initial impulses (that depend on the effective radii and their orientation) of the secondary CVRs are unknown.
One special case allows an exact analysis of the consecutive trajectory of one of the electrons -- when a secondary CVR has a small initial radius ($R_0 \ll R$). Then its initial deformation and direction of motion can be disregarded, because, under the pull of the electric field, it quickly gains sufficient energy and impulse along the $x$-direction, so to a good accuracy can be treated as a circular vortex ring \cite{SamuelsDonnelly1991,CVRTsubota}. If such singly-charged loops are created after collisions at some $x=x_1$, their arrival at the collector ($x=d$) at time $\tau_2(x_1) = \tau_1\left[\left(\frac{x_1}{d}\right)^2+\left(1-\frac{x_1}{d}\right)^2\right]$ will be earlier than of any other CVRs with either larger initial size (slower) or double charge (more energetic, hence, slower).
%[The uncharged small vortex loops will, of course, be the fastest, however, they do not contribute to the collector current and are hence undetected.]
The earliest arrival time,
\begin{equation}
\tau_2^* \equiv \min(\tau_2)= \frac{\tau_1}{2} ,
\label{tau2}
\end{equation}
will be for collisions at $x_1=d/2$.
Furthermore, if a secondary small ring grows and then reconnects with another vortex loop at some point $x=x_2$ ($x_1<x_2<d$), and this creates a new small singly-charged ring, the latter will arrive at the collector at time $\tau_3(x_1,x_2) = \tau_1\left[\left(\frac{x_1}{d}\right)^2 +\left(\frac{x_2}{d}-\frac{x_1}{d}\right)^2 +\left(1-\frac{x_2}{d}\right)^2\right]$. The earliest arrival, at time
\begin{equation}
\tau_3^* \equiv \min(\tau_3)= \frac{\tau_1}{3} ,
\label{tau3}
\end{equation}
of these second-generation secondary CVRs will correspond to two reconnections at $x_1=d/3$ and $x_2=2d/3$.
%This early-time cut-offs at $t=\tau_1/2$ were indeed observed experimentally (Figs.\,1\&2). This confirms that the broadening of $I_{\rm c}(t)$ is direct evidence for reconnections of vortex rings following their binary collisions.
% Let us assume that upon a collision of two primary CVRs of radii $R(x_*)$, at $x=x_*$ ($0 \leq x_* \leq d$) and time $\tau_* = \left(\frac{x_*}{d}\right)^2\tau_1$ (Eq.\,\ref{tau1} with approximation $R_0 = 0$) they are replaced by two secondary CVRs with the $x-$components of impulse $\alpha R(x_*)$ and $\beta R(x_*)$. The values of $\alpha$ and $\beta$ (let $0 < \alpha <1$, $1 < \beta <\sqrt{2}$) depend on the impact parameter. (No further collisions are assumed in this model, which is obviously only applicable at small densities $n$.) The arrival of secondary charged rings at the collector will produce the broadened current signal: the larger rings will now arrive later than $t=\tau_1$, i.\,e. contributing to the late-time broadening or ``tail'' of the collector current $I_c(t)$. On the other hand, the smaller rings will arrive earlier, at time $\tau_2 (x_*, \alpha)$. It is impossible to calculate the arrival time of secondary rings of arbitrary sizes, $R_2 \sim R(x^*)$, because we do not know the fraction of energy disspated (including that converted into Kelvin waves and lateral motion). We can only quote the longest possible time for the arrival of the larger secondary ring, assuming that it absorbed the total initial impulse of the two primary rings, $2 I(R(x^*))$, and both their electrons, $\tau_2(2e) = $. The experiment is consistent with such long-time broadenning (tail), yet no sharp long-time cut-off is observed; this is due to the fact that these larger (slower) rings absorb further CVRs from behand and thus become seeds of vortex tangles (QT), see below.
%However, there is one special case which can be treated accurately: when very small ($R_0 \ll R(d)$) charged rings are produced as a result of reconnections. The magnitude and direction of the initial impulase of such a ring can be neglected; it will quickly gain a much larger impulse in $x-$direction. The shortest arrival time is for singly-charged rings $\alpha \rightarrow 0$: $\tau_2 = \tau_* + \left(\frac{d-x_*}{d}\right)^2\tau_1$, and its minimum is at $x_* = d/2$:
%{\bf Experimental details.}
The experimental cell \cite{WalmsleyJLTPPobell}, a cube-shaped volume of side $d=4.5$\,cm, was filled with isotopically-pure liquid $^4$He \cite{pure4He} at pressure 0.1\,bar and temperature 0.05\,K (see inset in Fig.\,1).
Seed CVRs were injected through a gridded opening in the center of one plate. They then traveled along the axis of the container ($x$-axis) towards the center of the opposite plate to the collector electrode, placed behind a Frisch grid of radius $r = 6.5$\,mm and geometric transparency $\theta=0.92$.
%The collector current was recorded after a current-voltage converter with response time of 0.03\,s.
All currents and potentials are quoted with the opposite sign as if electrons had {\it positive} charge $e$. CVRs were subject to the propelling field set by potentials of plates $\phi(0)=0$ and $\phi(d)=U$, thus gaining energy $eU$ while travelling between the injector and collector grids \cite{CommentBuoyantRings}. The dependence $\phi(x)$ was close to the linear $\phi = U\frac{x}{d}$ (see inset in Fig.\,2).
The seed CVRs resulted from reconnections within the dense vortex tangle, generated by the current of electrons emitted from a tungsten tip \cite{GolovIshimoto} behind the injector grid through the voltage $U_{\rm tip}$.
These seed CVRs are injected in a broad range of angles; however, the impulse gained from the strong driving field quickly forces them to travel in nearly the same $x$-direction with a relatively narrow distribution of radii.
%Injected uncharged vortex rings do not grow and hence leave the vicinity of the beam of CVRs very quickly (and they are not detected by the collector), while multiply-charged seed rings are slower and hence trail behind the primary singly-charged CVRs.
The intensity and duration of the injected pulse were controlled by adjusting $U_{\rm tip}$ and its duration $\Delta t$ (all data presented here are for $\Delta t = 0.2$\,s).
%Most emitted electrons terminated at the injector grid but some penetrated it, riding on the seed CVRs.
For the same $U_{\rm tip}$ and $\Delta t$, the total charge injected through the grid was increasing with increasing drive voltage $U$ nearly linearly for all studied voltages. To quantify the time of flight $\tau_1$ of CVRs, we take the time interval between the middle of the tip voltage pulse and the position of the maximum of $I_{\rm c}(t)$ (and subtract the electronics response time of 0.03\,s).
% Due to the lag between the tip voltage and the peak intensity of the injection of seed CVRs caused by the intrinsic dynamics of the charged tangle, this procedure systematically overestimates $\tau$ by as much as $\sim 0.2$\,s; this translates into an overestimation of the initial radius of the seed CVRs $\bar{R_0} \approx 0.5$\,$\mu$m by as much as a factor of two. Yet, these uncertainties are insignificant for experiments with high $U>100$\,V (so that $R(d)\sim 5$\,$\mu$m $\gg R_0$), and do not affect our main observations of the changes in the shape and amplitude of the collector peaks.
%{\bf Observations.}
\begin{figure}
\includegraphics[width=7cm]{FigIvsTvarUtip2.eps}
\caption{(color online) Records of collector current, all for the same drive voltage $U=135$\,V, but different tip voltages $U_{\rm tip}$. The theoretical arrival time for primary CVRs, $\tau_1$ (Eq.\,\ref{tau1}), and of the earliest arriving secondary CVRs of first generation, $\tau_2^*$ (Eq.\,\ref{tau2}), and of second generation, $\tau_3^*$ (Eq.\,\ref{tau3}), are shown by arrows. Inset: Experimental cell with electric field pattern.}
\label{FigIvsTvarUtip2}
\end{figure}
Typical records of the collector current, $I_{\rm c}(t)$, following the injection of a pulse of CVRs are shown in Fig.\,\ref{FigIvsTvarUtip2}. These are all for the same drive voltage $U=135$\,V but several different injection currents. There is a well-defined peak at time $\tau_1 \approx 1.0$\,s corresponding to the arrival of primary CVRs. With increasing density of CVRs, this peak initially grows in magnitude while maintaining its position, $\tau_1$, and width, $\sim \Delta t$. At higher numbers of injected CVRs, however, a broad pedestal begins to grow, coexisting with the original peak (whose magnitude is now saturated). This broad peak-pedestal is due to the secondary CVRs that result from collisions between primary CVRs.
The current at $t > \tau_1$ reflects the arrivals of larger secondary vortex loops, while
%including the long ``tail'' consisting of multiply-charged vortex loops from multiple collisions.
% -- which constitute the precursor of the slow tangle of charged vortex lines, observed at higher densities of injected CVRs.
that at the earlier arrival times $t < \tau_1$ are from smaller secondary CVRs.
A sharp step builds up at $\tau_2^*=\tau_1/2$, coinciding with the earliest possible arrival of the first generation of secondary CVRs (Eq.\,\ref{tau2}). At the highest intensity of injection, another sharp step begins to form at $\tau_3^*=\tau_1/3$, corresponding to the earliest possible arrival of the second generation of secondary CVRs (Eq.\,\ref{tau3}).
\begin{figure}
\includegraphics[width=7cm]{FigIvsTvarU.eps}
\caption{
(color online) Records of collector current, all for the same $U_{\rm tip}=440$\,V, but different values of $U$. The right, middle and left arrows of the corresponding color point at the arrival time of primary CVRs, $\tau_1$, and the theoretical earliest arrival times for secondary CVRs of first (Eq.\,\ref{tau2}) and second (Eq.\,\ref{tau3}) generation, respectively.
The values of $\tau_1$ are defined as the positions of sharp peaks, except for $U=270$\,V -- where the peak due to primary CVRs is swamped by the broad pedestal due to secondary CVRs, hence the theoretical value of $\tau_1$ (see Fig.\,3) is used. For $U=22.5$\,V and 67.5\,V, the signal is too faint for the step at $\tau_3^*$ to be observable. Inset: The electrostatic potential $\phi(x)$ along the cell's axis.}
\label{FigIvsTvarU}
\end{figure}
In Fig.\,\ref{FigIvsTvarU}, we show $I_{\rm c}(t)$, similar to those in Fig.\,\ref{FigIvsTvarUtip2}, but now for the same tip voltage $U_{\rm tip} = 440$\,V and four different drive voltages $U$. With increasing $U$, the position of the peak $\tau_1$ increases as expected for isolated CVRs, Eq.\,\ref{tau1}. The peak's magnitude $I_{\rm m}(U)$ initially grows with $U$ but then, above $U=68$\,V, decreases -- even though the total collected charge $Q_{\rm c} = \int_0^\infty I_{\rm c}(t) \dif t$ keeps increasing. Simultaneously, the broad pedestal due to secondary CVRs progressively overgrows the primary peak until completely swamping it at $U=270$\,V. The sharp steps due to the earliest possible arrivals of secondary CVRs of first generation at $\tau_1/2$ (Eq.\,\ref{tau2}) and second generation at $\tau_1/3$ (Eq.\,\ref{tau3}) are labeled by arrows.
%As one can see from Fig1--2, sharp short-time cut-offs of the collector current, at times $\tau_1/3$ and $\tau_1/2$ have been observed indeed.
We thus obtained quantitative evidences of either single or two consecutive reconnections of CVRs during their motion from the injector to collector. Furthermore, the substantial contribution to the collector current right after the cut-offs indicates that very small CVRs are created with high probability. This might contradict the expectations that reconnections result in vortex loops of size {\it comparable} to the radius of curvature of the initial vortex lines \cite{BarenghiRingEmission2002,KS}, but would support the picture of a cascade of small vortex rings created by large-amplitude Kelvin waves generated after a reconnection of nearly antiparallel vortex lines when dissipation is small \cite{Svistunov1995,Kerr2011,Kursa2011}.
%The question whether trapped ions might help to pinch-off small CVRs requires further research.
\begin{figure}
\includegraphics[width=7cm]{FigTvsU.eps}
\caption{(color online) The elapsed times between the middle of the emitter pulse and maximum of $I_{\rm c} (t)$ vs. $U$. Closed symbols correspond to resolvable sharp peaks due to primary CVRs; open symbols are for the broadened peaks due to secondary CVRs where sharp peaks due to primary CVRs are no longer visible. Vertical bars indicate the peak width at half-maximum. The solid line shows the theoretical arrival times $\tau_1$ (Eqs.\,1--2) for $R_0=0.5$\,$\mu$m. The dashed line shows the approximate solution for $\Lambda = 13$ and effective $R_0=0.8$\,$\mu$m (Eq.\,\ref{tau1}). Cartoons illustrate the expansion and progression of: (left) isolated primary (red) CVRs at low density, and (right) a reconnection resulting in secondary (blue) CVRs.}
\label{FigTvsU}
\end{figure}
In Fig.\,\ref{FigTvsU}, the experimental arrival times $\tau_1(U)$ for several intensities of injection are plotted. For small drive voltages $U$ (i.\,e. when the radii of CVRs $R(d)$ and density of CVRs $n(d)$ are small) the experimental points agree with the theory for isolated CVRs (from Eqs.\,\ref{vE}--\ref{R_appr}). To characterize the range of the distribution of times of flight, the vertical bars show the width of $I_{\rm c}(t)$ at the $0.5 I_{\rm m}$ level. One can see that at small $U$ the width is constant, being equal to the injection duration $\Delta t = 0.2$\,s. Above a certain value of $U \sim 100$\,V (that decreases with increasing injection intensity, $U_{\rm tip}$), the collector pulse broadens and the position of the maximum of $I_{\rm c}(t)$ no longer agrees with the theoretical prediction for isolated CVRs (this coincides with the complete disappearance of the sharper peak due to primary CVRs, as on the trace for $U=270$\,V in Fig.\,\ref{FigIvsTvarU}); secondary CVRs dominate $I_{\rm c}(t)$ in these conditions of high $n$ and $R$.
% (it seems, as, in the presense of electric field, an electron always seats in a self-made sharp casp of vortex line, this casp might be a natural seed for the small CVR to be pinched off [Kelvin waves, generated upon a reconnection, might grow strongly-non-linear when approaching this anchor point]. However, this mechanism will probably suggest that all electyrons should be shaken-off upon any reconnection -- which is clearly not the case. Data at different temperatures might help clear out this question. Also, we should try analytical/numerical modelling of populations of small CVRs subject to different models.) If only about one small secondary CVRs can be created upon a collision of two primary CVRs, and the probability of the transfer of two trapped electrons to secondary rings is proportional to their lengths, one would expect the current delivered by the secondary CVRs of nearly the smalles tpossible size to be too small to be observable. Yet, while approaching the cut-off value of $t=\min{\tau_2}$, the contributon to $I_{\rm c}(t)$ was still quite comparable with the typical values of $I_{\rm c}$ for larger secondary rings (at larger $t$). Unless the process of pinching-off of a secondary CVR can be facilitated by the locally present trapped electron (this will be checked in our further experiments and numerical simulations), we should conclude that many small secondary vortex rings are produced afettr a typical collision of primary CVRs.
%{\bf Density of primary vortex rings.}
% \begin{figure}
%\includegraphics[width=7cm]{FigImaxQtotalU.eps}
%\caption{(color online) Magnitudes of peaks of collector current vs. driving voltage $U$ for several different intensities of injection ($U_{\rm tip}$). $\Delta t = 0.2$\,s.
%The solid line indicates the dependence $I_{\rm m} \propto U^{-2}$.
% Asterisks (right vertical axis) show the total collected charge $Q_{\rm c}$ for $U_{\rm tip} = 360$\,V; the dotted line guides the eye.}
%\label{FigImaxQtotalU}
%\end{figure}
\begin{figure}
\includegraphics[width=7cm]{FigNvsR.eps}
\caption{(color online) Density of primary CVRs $n$ vs. their radius $R$ at $x=d$ measured for several different drive voltages $U$ (top axis) and injector tip voltages, $U_{\rm tip}$ (see legend).
%The values of $n$ are calculated from the magnitudes of peaks of collector current $I_{\rm}$ using Eq.\,.. while $R$ are obtained from the peaks' positions (arrival times) $\tau_1$ via Eqs.\,\ref{R_appr,tau1}.
%Dashed line shows the dependence $n \propto R^{-1}$ (Eq.\,\ref{nR2}).
}
\label{FigNvsR}
\end{figure}
The density of primary rings reaching the collector, $n_1 \equiv n(d)$, can be found from the value of the collector current at its maximum (but only for the records $I_{\rm c}(t)$ that have a sharp peak at $t=\tau_1$, dominant over the pedestal due to the secondary CVRs),
\begin{equation}
I_{\rm m} \approx \theta \pi r^2 e n_1 v_1 \approx \frac{1}{8}\theta r^2\rho\kappa^3(\Lambda-1)^2\frac{n_1}{U},
\label{Im}
\end{equation}
where the relation $v_1 \equiv v(d) \approx \frac{\rho\kappa^3(\Lambda-1)^2}{8\pi eU}$ (from Eqs.\,\ref{vE}--\ref{R_appr}) has been used.
In Fig.\,\ref{FigNvsR}, we plot $n_1$ (calculated from the experimental values of $I_{\rm m}(U)$ using Eq.\,\ref{Im}) vs. $R_1 \equiv R(d)$, the radii of primary CVRs near the collector, (calculated from Eq.\,\ref{R_appr} with $R_0=0$).
Again, there are two regimes: at low $R_1$, $n_1$ increases with an increase of either $U_{\rm tip}$ or $U$, but at large $R_1$, $n_1$ becomes a {\it decreasing} function of $U$, independent of $U_{\rm tip}$. Yet, the total injected charge (as measured by the integral of the collector current) increases monotonically at all $U$.
I.\,e., only at small $R_1$ and $n_1$, the CVR's density can be controlled by varying the injection current. At high $R_1$, only a small fraction of charge arrives with primary CVRs of now {\it universal} density $n_1 \approx 500{\rm \,cm}^{-2} R_1^{-1}$; the rest (secondary CVRs) contribute to the broad pedestal in $I_{\rm c}(t)$ (Fig.\,\ref{FigIvsTvarUtip2}--\ref{FigIvsTvarU}).
% At high densities of injected CVRs, the beam at $x=d$ was much wider than the diameter of the collector grid. We hence can use the collector's area $A=\pi r^2$ and known velocity of isolated CVRs $v(d)$ to evaluate the peak number density of primary CVRs, $n(d)$, from the experimental value of $I_{\rm m}$:
The density of primary rings in the beam $n(x,t)$, along the trajectory $x(t)$, evolves according to Eq.\,11 in \cite{Supplement}:
\begin{equation}
\frac{\dif n}{\dif t} = - n\frac{\dif v}{\dif x} - f,
\label{RateEquation}
\end{equation}
where $-f$ is the rate of losses, per unit volume and time, due to ring-ring collisions. For small injected density and radii, the collisions can be neglected, $f=0$, and the solution for the density near collector is
\begin{equation}
n_1 = \frac{8\pi I_m U}{\theta\pi r^2 \rho \kappa^3 (\Lambda -1)^2},
\label{n1}
\end{equation}
i.\,e. it can be varied by changing either the injected density of CVRs (characterized by $I_m$) or drive voltage $U$ -- as observed in the experiment. When collisions become rife at higher densities and radii, accounting for the removal of primary CVRs due to their binary collisions results in
\begin{equation}
n_1R_1 = \frac{3}{\sigma_1 ' \delta \! R d} = 4\times 10^3{\rm\,cm}^{-2},
\label{nR1}
\end{equation}
where we used $\delta \! R = 0.5$\,$\mu$m and the geometric cross-section for collisions $\sigma_1 = \sigma_1 ' R^2$ with $\sigma_1 ' = 4\pi$. Furthermore, the subsequent removal of primary CVRs that bump into the slower loops \cite{CommentPile-ups} left after the described collisions of primary CVRs will result in the solution,
\begin{equation}
n_1R_1=\left( \frac{40\pi}{(\Lambda-1/2)\kappa \sigma_1 ' \sigma_2 ' \delta \! R \Delta t d} \right)^{1/2} = 1\times 10^3{\rm\,cm}^{-2},
\label{nR2}
\end{equation}
where the geometric cross-section for these pile-ups is estimated as $\sigma_2 = \sigma_2 ' R^2$ with $\sigma_2 ' = 4\sqrt{2}\pi$.
Both Eq.\,\ref{nR1} and Eq.\,\ref{nR2} reproduce the experimental universal dependence $n_1R_1 = 500$\,cm$^{-2}$ qualitatively, while Eq.\,\ref{nR2} is actually quite close {\it quantitatively} (our 1-dimensional model underestimates $f$ by disregarding the transverse component of the relative motion of primary CVRs and hence overestimates the value of $n_1R_1$ by a factor of 2 or so \cite{Supplement}).
The fact of occasional piling-up of many primary CVRs, in turn, helps to explain the appearance of large-scale velocity fluctuations in the ensuing vortex tangle.
In the initial random beam of primary CVRs, the fluctuations of the coarse-grained velocity on the length scales greater than $\sim n^{-1/3}$ (``quasi-classical'' flow) are small; the energy spectrum is concentrated around the small length scale of order $R$.
%Below that scale, the tangle can be treated as a random tangle (``ultra-quantum turbulence'', although strongly polarized because the beam is unidirectional).
The small secondary vortex rings, observed in this work, and Kelvin waves excited by reconnections are evidence of the direct cascade of energy towards smaller length scales \cite{Svistunov1995, Tsubota2000, BarenghiRingEmission2002, NazarenkoReconnection, KS, Nemirovskii2010}.
However, following any of the pile-ups of many vortex rings, strong fluctuations of the coarse-grained velocity field on the ``quasi-classical'' length scales $\gg n^{-1/3}$ are being created.
This is the inverse cascade of energy in this strongly anisotropic system \cite{Yamamoto2012, Baggaley2013}.
% Unlike the mechanism relying on just merger of two rings (and hence increasing their local velocity scale by 40\%) but not going beyond $\sim n^{-1/3}$ discussed in \cite{Barenghi}, this one relyes on avalanche-like catastrophic expansion of the tangled loops, and hence, is a very different process.
%Another mechanism, which works in parallel, is due to the fact that the beam does not occupy all space but has a finite cross-sectoin, surrounded by zero mean velocity field. Hence, the impulse is only concentrated within the beam; any tangling of the vortex rings comprising the beam results in spatial modulations of the coarse-grained density of impulse on length scales of order the beam's cross-diameter.
To summarize, we obtained quantitative evidence for collisions and reconnections of pairs of unidirectional vortex rings of similar radii that result in the creation of vortex loops of unequal size, including many small ones. We observed discrete steps at the time dependence of the collector current, that correspond to the earliest arrivals of the first and second generations of secondary CVRs. As each collision can cause a removal of many primary vortex rings, increasing the density of injected CVRs results in a new state in which the density of primary vortex rings is maintained at the critically depleted level independent of their initial density. The larger loops produced in the collisions become the seeds of quasi-classical QT with large-scale flow structures, which appear out of a seemingly random beam of small quantized vortex rings.
%: the {\it inverse cascade} of energy \cite{Yamamoto2012, Baggaley2013} in action.
% The unprecedented level of control of the properties (energy, velocity, orientation, radii, density) of primary vortex rings, never achieved in previous experiments, allowed us to observe discrete steps at the time dependence of the collector current, which correspond to the earliest arrivals of the second and third generations of charged vortex rings.
This work was supported by the Engineering and Physical Sciences Research Council (Grants No. GR/R94855, EP/H04762X/1 and EP/I003738/1).
\section{Supplementary material: density of unidirectional vortex rings subject to reconnections}
\subsection{Introduction}
In these supplementary notes, we describe effects of collisions of unidirectional charged vortex rings (CVRs) due to the small variations, $\sim \delv$, of their velocities around the mean value $v$. Any such collision of primary CVRs generally results in secondary CVRs of quite different size -- hence depleting the number of the primary CVRs.
%Consider a beam of rings of mean radius $R$ and density $n$ with a small random spread $\delta R \sim \bar{R_0}$, all moving in the $x$-direction with mean velocity $v \sim \kappa /R$.
We consider purely one-dimensional trajectories of CVRs, $x(t)$, between the injector at $x=0$ and collector at $x=d$ \cite{1-d}. We assume a uniform electric field $U/d$, a negligible initial radius of CVRs, $R_0 \rightarrow 0$, and a constant logarithmic parameter $\Lambda \equiv \ln \frac{8R}{a_0} \approx 13$ (where $a_0 \approx 1.3$\,\AA\ is the core radius). Within the time of flight from the injector to collector, $\tau_1$, the lengthening of the beam, caused by the range of velocities $\delta v$, does not exceed the duration of injection $\Delta t = 0.2$\,s and is, hence, neglected.
The velocity and energy of a quantized vortex ring of radius $R$ carrying one electron of charge $e$:
\begin{equation}v = \frac{\kappa}{4\pi R}\left(\Lambda-\frac{1}{2}\right),\label{v(R)}\end{equation}
\begin{equation}{\cal E} = \frac{\kappa^2 \rho R}{2}\left(\Lambda-\frac{3}{2}\right) = e\frac{U}{d}x,\label{E(R)}\end{equation}
where $\rho=0.145$\,g\,cm$^{-3}$ is the density of superfluid helium and $\kappa = 1.0\times 10^{-3}$\,cm$^2$\,s$^{-1}$ is the quantum of circulation.
%In what follows, we assume $\Lambda = const$.
%The time of travel between any two points is {equation}\tau = \int_a^b \! v^{-1} \dx \approx \frac{4\pi}{\kappa(\Lambda - 1/2)} \int_a^b \! R \dx{equation} (i.\,e. proportional to the spatial average of the radius $R$).
The radius grows linearly with $x$,
\begin{equation}R = R_1\frac{x}{d},\label{R(x)}\end{equation}
where the radius of CVRs at collector, $x=d$, is
\begin{equation}R_1 = \frac{2eU}{\rho\kappa^2(\Lambda-3/2)}\label{R1}.\end{equation}
The time of flight between $x=0$ and $x=x_1$ is
\begin{equation}t(x) = \int_0^{x} v^{-1}dx' = \tau_1\frac{x^2}{d^2},\label{t(x)}\end{equation}
i.\,e. the trajectory of primary CVRs is
\begin{equation}x(t) = d\left(\frac{t}{\tau_1}\right)^{1/2},\label{x(t)}\end{equation}
where the time of flight from injector to collector is
\begin{equation}\tau_1 = \frac{4\pi e d}{\rho\kappa^3(\Lambda-1)^2}U = \frac{2\pi d}{\kappa(\Lambda-1/2)}R_1.\label{tau1}\end{equation}
%From now on, we use approximarion $R_0 = 0$.
The velocity field, its gradient and pulse length in $x$-space can be conveniently expressed (including the explicit time dependence along the trajectory $x(t)$):
\begin{equation}v\equiv \frac{\dif x}{\dif t} =\frac{d^2}{2\tau_1}x^{-1} = \frac{d}{2}\tau_1^{-1/2}t^{-1/2},\label{v(x)}\end{equation}
%(velocity at collector, for instance, $v_d = \frac{d}{2\tau_1}$), hence,
\begin{equation}\nabla v = -\frac{d^2}{2\tau_1} x^{-2} = -\frac{1}{2}t^{-1},\label{nablav}\end{equation}
%and {equation}\dif v = -\frac{d}{4\sqrt{\tau_1}}t^{-3/2} \dif t{equation} (along the trajectory).
\begin{equation}\Delta x = v\Delta t = \frac{d^2\Delta t}{2\tau_1}x^{-1} = \frac{d\Delta t}{2}\tau_1^{-1/2}t^{-1/2}.\label{Deltax}\end{equation}
%------------
\subsection{Evolution of the density of CVRs}
The number density of primary rings in the beam $n(x,t)$, along the trajectory $x(t)$, evolves according to:
\begin{equation}
\frac{\dif n}{\dif t} \equiv \frac{\partial n}{\partial t} + v \frac{\partial n}{\partial x} = \left(-\frac{\partial (nv)}{\partial x} -f \right) + v \frac{\partial n}{\partial x} = - n \frac{\dif v}{\dif x} - f,
\label{RateEquation}
\end{equation}
where $-f$ is the rate of losses due to ring-ring collisions.
%or {equation}\frac{\dif n}{\dif x^2} + \frac{\kappa(\Lambda-1/2) (\sigma R^{-2})\delR \tau_1}{8\pi d^2} n^2 - \frac{x^{-2}}{2}n = 0.{equation}
For small $n$, at which $f\rightarrow 0$, the number of CVRs is conserved, and the solution is
$nv = const$, i.\,e. $n = A x$.
%In this regime, the density $n$ can be varied by changing either the intensity of injection or the driving voltage $U$.
% holds for the product $AB \ll x \sim d$, i.\,e. for either small injected density of CVRs (small $I_m$) or small $U$ or $\delR$.
Experimentally, the value of constant $A$ can be determined from the magnitude of the peak of the collector current due to primary CVRs,
\begin{equation}I_m = \theta\pi r^2 e n_1 v_1 = \theta\pi r^2 e A d^2 (2\tau_1)^{-1},\label{Im}\end{equation}
where $r$ and $\theta$ are the radius and transparency of the collector grid, $n_1\equiv n(d)$ and $v_1\equiv v(d)$.
We arrive at
\begin{equation}
n(x) = \frac{2\tau_1 I_m}{\theta\pi r^2 e d^2} x = \frac{8\pi I_m U}{\theta\pi r^2 \rho \kappa^3 (\Lambda -1)^2d} x.
\label{n1}
\end{equation}
The density at collector is
\begin{equation}
n_1 = \frac{8\pi I_m U}{\theta\pi r^2 \rho \kappa^3 (\Lambda -1)^2}.
\label{n1}
\end{equation}
Thus, in this regime of conserved primary CVRs, their density can be varied by changing either the intensity of injection (as measured by $I_m$) or drive voltage $U$.
\subsection{Collisions of primary CVRs}
%\section{Frequency of reconnections of primary CVRs}
%Primary CRVRs reconnect with themselves at rate $f_1$ per unit time in unit volume.
%) and with slowly-moving vortex lines (either larger vortex loops left behind after binart reconnections of primary CVRs or remanent vortex lines from previous flows) at rate $f_2$.
% Hence, the rate of losses of primary CVRs is $f=f_1+f_2$.
For each CVR, the probability of a collision with another CVR per unit time is $n\sigma_1 \delv $, where $\delv = v\delR/R$ (as $v \propto R^{-1}$), and the cross-section is $\sigma = \sigma_1 ' R^2$, where $\sigma_1 ' \sim 1$. Naive geometric considerations suggest $\sigma_1 ' = 4\pi $, although further research into this problem is necessary \cite{Paul_sigma}.
If each collision effectively removes two primary CVRs from the coherent beam, the total number of removed CVRs per unit time and unit volume:
\begin{equation}f_a = n^2 \sigma_1 \delv = \frac{\Lambda-1/2}{4\pi}\kappa \sigma_1 ' \delR n^2.\label{f1}\end{equation}
%where the factor $\frac{1}{2}$ is to avoid double-counting, and $\sigma R^{-2} \sim 1$ is typically a constant of order unity.
%If we rewrite it as
%{equation}\frac{\dif n}{\dif x^2} + 2B n^2 - \frac{1}{2}x^{-2}n = 0,{equation}
%{equation}\frac{\dif n}{\dif t} + \frac{3B}{2} n^2 - \frac{n}{2t} = 0,{equation}
After using Eqs.\,\ref{nablav}\&\ref{tau1}, and $f=f_a$ from the above expression, Eq.\,\ref{RateEquation} becomes:
%{equation}\frac{\dif n}{\dif t} = -\frac{\kappa(\Lambda-1/2) (\sigma R^{-2}) \delR}{4\pi} n^2 + \frac{n}{2t}.{equation}
\begin{equation}\frac{\dif n}{\dif t} - \frac{n}{2t} + \frac{d \sigma_1 ' \delR R_1}{2\tau_1} n^2 = 0.\label{RateEquation1}\end{equation}
Its asymptotic solution is
\begin{equation}n = \frac{3}{\sigma_1 ' \delR d}R_1^{-1}\frac{\tau_1}{t}.\label{Solution1}\end{equation}
%{equation}B = \frac{(\Lambda - 1/2)\kappa (\sigma R^{-2})\delR \tau_1}{16\pi d^2} = \frac{e (\sigma R^{-2})\delR U}{4 (\Lambda-3/2) \rho\kappa^2 d}.{equation}
At $t=\tau_1$,
%{equation}n(d) = \frac{4 (\Lambda-3/2) \rho\kappa^2 }{e (\sigma R^{-2})d \delR U},{equation}
\begin{equation}
n_1R_1 = \frac{3}{\sigma_1 ' \delR d}.
\label{nR1}
\end{equation}
The RHS is independent of both the injection intensity and $U$. For $\sigma ' =4\pi$ and $\delR = 0.5$\,$\mu$m, it is equal to $4\times10^3$\,cm$^{-2}$, which is a factor of 8 greater than the experimental value of $5\times10^2$\,cm$^{-2}$. There might be several reasons for such a discrepancy. Firstly, the cross-section for the effective removal of CVRs from the coherent beam of similar orientation and velocities might be several times greater than the geometric guess $\sigma = 4\pi R^2$. Secondly, in this model we restricted all CVRs to motion along the $x$-axis with only a spread of radii; in reality they also have a small random spread of directions of motion caused by the conserved transverse component of the impulse of the initial CVRs -- our estimates show that its contribution to the frequency of collisions is comparable with that calculated here.
%Finally, a different mechanism might help limit the product $n_1R_1$ -- which is described in the next section.
%--------------
\subsection{Multiple pile-ups of primary CVRs}
Another process that should further limit the density of primary CVRs is the removal of all CVRs that bump into the secondary vortex rings formed upon any binary collision discussed above.
Large secondary CVRs appear with each primary collisions at a rate, per unit volume and time,
\begin{equation}\frac{f_a}{2} = \frac{(\Lambda-1/2)\kappa \sigma_1 ' \delR}{8\pi} n^2. \label{f12}\end{equation}
%, that can be somehow smaller than $\sigma$ (which also accounted for close encounters of two CVRs but without reconnections) from the previous section.
As these larger rings are considerably slower than the primary ones, they block all the primary CVRs from behind within the effective cross-section $\sigma_2 = \sigma_2 ' R^2$, where $\sigma_2 ' \sim 1$, and the upper limit on its geometrical value is $\sigma_2 ' = 4\sqrt{2}\pi \approx 18$ \cite{sigma2}. The typical number of primary CVRs lost in such a capture (the ``multiplication factor'') is $N_2 \sim n\sigma_2\frac{\Delta x}{2}$. The rate of losses per unit volume and time is hence
%{equation}f_2 = N_2\frac{f_1}{2} = \frac{(\Lambda-1/2)a \kappa(\sigma_1 R^{-2}) \delR}{16\pi} R^2 \Delta x n^3 ={equation}
\begin{equation}f_b = N_2\frac{f_a}{2} = \frac{(\Lambda-1/2) \kappa \sigma_1 ' \sigma_2 ' \delR R_1^2 \Delta t d}{32\pi \tau_1^{3/2}} t^{1/2} n^3.\label{f2}\end{equation}
With $f=f_b$, Eq.\,\ref{RateEquation} becomes
%{equation}b = \frac{(\Lambda-1/2)a \kappa(\sigma_1 R^{-2}) \delR R_1^2 \Delta t d}{32\pi \tau_1^{3/2}},{equation}
%this results in the rate equation for $n$,
%{equation}\frac{\dif n}{\dif t} - \frac{n}{2t} + bt^{1/2} n^3.{equation}
\begin{equation}
\frac{\dif n}{\dif t} - \frac{n}{2t} + \frac{(\Lambda-1/2) \kappa \sigma_1 ' \sigma_2 ' \delR R_1^2 \Delta t d}{32\pi \tau_1^{3/2}} t^{1/2} n^3 =0.
\label{RateEquation2}\end{equation}
Its asymptotic solution is:
\begin{equation} n = \left( \frac{40\pi }{(\Lambda-1/2) \kappa \sigma_1 ' \sigma_2 ' \delR \Delta t d} \right)^{1/2} R_1^{-1} \left(\frac{\tau_1}{t}\right)^{3/4}.\label{Solution2}\end{equation}
%{equation}n=\sqrt{\frac{5}{4b}}t^{-3/4}.{equation}
For $t=\tau_1$ this gives
\begin{equation}
n_1R_1=\left( \frac{40\pi}{(\Lambda-1/2)\kappa \sigma_1 ' \sigma_2 ' \delR \Delta t d} \right)^{1/2},
\label{nR2}
\end{equation}
independent of the injection intensity and $U$. For $\sigma_1 ' = 4\pi$, $\sigma_2 ' = 4\sqrt{2}\pi$, $\delR = 0.5$\,$\mu$m, $\Delta t = 0.2$\,s and $d=4.5$\,cm, this produces $n_1R_1 = 1.0\times 10^3$\,cm$^{-2}$, which is only factor of 2 greater than the experimental value. Furthermore, as we only considered the effect of longitudinal variations of the velocities $\delv$ of primary CVRs but disregarded the comparable effect from their transverse velocities, the agreement is quite good.
%This might be possible if pile-ups of primary CVRs is an avalanche-like process, rapidly widening its effective capture cross-section $\sigma_2$ after every new reconnecting CVR due to the buildup of a vortex tangle. Then a critical state will be maintained at the density at which these catastrophic avalanches are not frequent.
\subsection{Conclusion}
To conclude, the solutions, $n_1R_1=const$, found for either of the simple models of removal of primary vortex rings upon collisions, Eq.\,\ref{nR1} and Eq.\,\ref{nR2}, qualitatively agree with our experimental observations. Moreover, the latter is actually in a reasonable quantitative agreement.
%The numerical values of the constant are not inconistent with the experimental findings, although more research into the appropriate values of $\sigma_1 ' = 4\pi$ and $\sigma_2 ' = 4\sqrt{2}\pi$
Thus multiple pile-ups explain the observed universal value of the product $n_1R_1$ when CVRs collide frequently due to their large numbers and large radii.
%It is likely that both processes contribute in parallel.
Further experiments with varyable flight path $d$, duration of injection $\Delta t$ and distribution of CVRs' radii $\delR$, as well as numerical simulations of the interaction between CVRs, will help improve our understanding of the microscopic processes that occur within beams of unidirectional vortex rings.
%\section{Cross-over between two regimes}
%As there are two limiting solutions of Eq.\,\ref{RateEquation}, one for small $R$ (no losses), Eq.\,\ref{n1}, and another for large $R$.
%For this universal assymptotic regime to be reached at $x\leq d$, the intensity of injection and applied voltage should be sufficiently strong. The value of collector current $I_m^*$, at which the cross-over from Eq.\,\ref{n1} to Eq.\,\ref{nR} happens is
%\begin{equation}I_m^* = \frac{3r^2\theta\rho\kappa^3(\Lambda-1)^2}{8(\sigma R^{-2})\delR d}U^{-1}.\label{Im*}\end{equation}
% Such a switch between two regimes is clearly seen in Figs.\,3--4 of main paper.
%===================================================
\begin{thebibliography}{99}
\bibitem{ShariffLeonard1992} K. Shariff, A. Leonard, Annu. Rev. Fluid Mech. {\bf 24}, 235 (1992).
% JOURNAL OF COMPUTATIONAL PHYSICS 37, 289-335 (1980) Review Vortex Methods for Flow Simulation A. LEONARD
%\bibitem{Karniadakis1993} G. E. Karniadakis and S. Orszag, Nodes, modes and flow codes, Physics Today, March, 34--42 (1993).
\bibitem{Donnelly} C. F. Barenghi, R. J. Donnelly, Fluid Dyn. Res. {\bf 41}, 051401 (2009).
\bibitem{KoplikLevine96} J. Koplik and H. Levine,
Phys. Rev. Lett. {\bf 76}, 4745 (1996).
\bibitem{Kivotides2003} D. Kivotides and A. Leonard,
Phys. Rev. Lett. {\bf 90}, 234503 (2003).
\bibitem{Kambe1971} T. Kambe and T. Takao, J. Phys. Soc. Japan {\bf 31}, 591 (1971).
\bibitem{Fohl1975} T. Fohl and J. S. Turner, Phys. Fluids {\bf 18}, 433 (1975).
\bibitem{Oshima1977} Y. Oshima and S. Asaka, J. Phys. Soc. Japan {\bf 42}, 708 (1977).
\bibitem{RR1964} G. W. Rayfield and F. Reif, Phys. Rev. {\bf 136}, A1194 (1964).
\bibitem{Kerr87} A. Pumir and R. M. Kerr,
Phys. Rev. Lett. {\bf 58}, 1636 (1987).
\bibitem{KoplikLevine93} J. Koplik and H. Levine, Phys. Rev. Lett. {\bf 71}, 1375 (1993).
\bibitem{Nore97} C. Nore, M. Abid, and M. E. Brachet,
Phys. Rev. Lett. {\bf 78}, 3896 (1997).
\bibitem{Kivotides2001} D. Kivotides, J. C. Vassilicos, D. C. Samuels, and C. F. Barenghi,
Phys. Rev. Lett. {\bf 86}, 3080 (2001).
\bibitem{Leadbeater2001} M. Leadbeater, T. Winiecki, D. C. Samuels, C. F. Barenghi, and C. S. Adams,
Phys. Rev. Lett. {\bf 86}, 1410 (2001).
% Sound Emission due to Superfluid Vortex Reconnections
\bibitem{Leadbeater2003} M. Leadbeater, D. C. Samuels, C. F. Barenghi, and C. S. Adams,
Phys. Rev. A {\bf 67}, 015601 (2003).
\bibitem{Chatelain2003} P. Chatelain, D. Kivotides, and A. Leonard,
Phys. Rev. Lett. {\bf 90}, 054501 (2003).
\bibitem{Mitani2006} A. Mitani and M. Tsubota,
Phys. Rev. B {\bf 74}, 024526 (2006).
\bibitem{Nemirovskii2006} S. K. Nemirovskii,
Phys. Rev. Lett. {\bf 96}, 015301 (2006).
\bibitem{Barenghi2006}
C. F. Barenghi, R. H{\"a}nninen, and M. Tsubota,
Phys. Rev. E {\bf 74}, 046303 (2006).
\bibitem{Alamari2008}
S. Z. Alamri, A. J. Youd, and C. F. Barenghi,
Phys. Rev. Lett. {\bf 101}, 215302 (2008).
\bibitem{Kerr2011} R. M. Kerr,
Phys. Rev. Lett. {\bf 106}, 224501 (2011).
\bibitem{Kursa2011} M. Kursa, K. Bajer, and T. Lipniacki,
Phys. Rev. B {\bf 83,} 014515 (2011).
\bibitem{Crow} S. C. Crow, AIAA J. {\bf 8}, 2172 (1970).
\bibitem{Hanninen2013} R. H{\"a}nninen,
Phys. Rev. B {\bf 88}, 054511 (2013).
\bibitem{Caplan2014} R. M. Caplan, J. D. Talley, R. Carretero-Gonzalez, and P. G. Kevrekidis,
arXiv:1404.4922v1 (2014).
\bibitem{Beweley2008} G. P. Beweley {\it et al.}, Proc. Natl. Acad. Sci. {\bf 105}, 13707 (2008).
\bibitem{Fonda2014} E. Fonda {\it et al.}, Proc. Natl. Acad. Sci. {\bf 111}, 4707 (2014).
\bibitem{Schwarz1988} K. W. Schwarz, Phys. Rev. B {\bf 38}, 2398 (1988).
\bibitem{VinenNiemela2002} W. F. Vinen and J. J. Niemela, J. Low Temp. Phys. {\bf 128}, 167 (2002).
\bibitem{Yano2013} Y. Nago, A. Nishijima, H. Kubo, T. Ogawa, K. Obara, H. Yano, O. Ishikawa, and T. Hata, Phys. Rev. B {\bf 87}, 024511 (2013).
\bibitem{Salman2013} H. Salman, Phys. Rev. Lett. {\bf 111}, 165301 (2013).
\bibitem{Svistunov1995} B. V. Svistunov, Phys. Rev. B {\bf 52}, 3647 (1995).
\bibitem{Tsubota2000} M. Tsubota, T. Araki, S. K. Nemirovskii, {\it Phys. Rev.}B {\bf 62} 11751 (2000).
\bibitem{BarenghiRingEmission2002} C. F. Barenghi and D. C. Samuels, Phys. Rev. Lett. {\bf 89}, 155302 (2002).
\bibitem{Nemirovskii2008} S. K. Nemirovskii Phys. Rev. B {\bf 77}, 214509 (2008).
\bibitem{Nemirovskii2010} S. K. Nemirovskii, Phys. Rev. B {\bf 81}, 064512 (2010).
\bibitem{KS} E. Kozik and B. Svistunov, Phys. Rev. B {\bf 77}, 060502(R) (2008).
\bibitem{Yamamoto2012} S. Yamamoto, M. Tsubota, and W. F. Vinen, J. Phys. Conf. Series, {\bf 400}, 012075 (2012).
\bibitem{Kondaurova2012} L. Kondaurova and S. K. Nemirovskii Phys. Rev. B {\bf 86}, 134506 (2012).
\bibitem{Baggaley2013} A. W. Baggaley, C. F. Barenghi, Y. A. Sergeev, Phys. Rev. B {\bf 85}, 060501 (2012); Phys. Rev. E {\bf 89}, 013002 (2014).
%\bibitem{VinenRings} Vinen Rings?
\bibitem{NazarenkoReconnection} S. Nazarenko, JETP Lett. {\bf 84}, 585 (2007).
% Kelvin wave turbulence generated by vortex reconnection
\bibitem{Bradley2005} D. I. Bradley, {\it et al.}, Phys. Rev. Lett. {\bf 95}, 035302 (2005).
\bibitem{PRLWalmsley2008} P. M. Walmsley and A. I. Golov, Phys. Rev. Lett. {\bf 100}, 245301 (2008).
\bibitem{TsepelinFluctuations} D. I. Bradley, {\it et al.}, Phys. Rev. Lett. {\bf 101}, 065302 (2008).
\bibitem{Roche} P.-E. Roche and C. F. Barenghi, Europhys. Lett. {\bf 81}, 36002 (2008).
\bibitem{WalmsleyLongInjection} A. I. Golov, P. M. Walmsley, J. Low Temp. Phys. {\bf 156}, 51 (2009).
\bibitem{WalmsleyPNAS2014} P. Walmsley, D. Zmeev, F. Pakpour, and A. Golov, Proc. Natl. Acad. Sci. {\bf 111}, 4691 (2014).
\bibitem{GlabersonDonnelly1986} W. I. Glaberson and R. J. Donnelly, {\it Structure, Distribution, and Dynamics of Vortices in Helium II}, edited by D. F. Brewer, Progress in Low Temperature Physics {\bf 9} (North-Holland, Amsterdam, 1986).
%\bibitem{CommentGamota} The contribution to the mean electric field due to the space charge of CVRs, that might reduce the width $\Delta x$ of the propagating cloud of CVRs \cite{GamotaPRL1971} was negligible in our case.
%\bibitem{GamotaPRL1971} G. Gamota, A. Hasegawa, C. M. Varma, Phys. Rev. Lett. {\bf 26}, 960 (1971).
\bibitem{SamuelsDonnelly1991} D. C. Samuels and R. J. Donnelly, Phys. Rev. Lett. {\bf 67}, 2505 (1991).
\bibitem{CVRTsubota} M. Tsubota and H. Adachi, J. Low Temp. Phys. {\bf 158}, 364 (2010).
% Dynamics of Charged Quantized Vortices, Ion-Vortex Complexes
%\bibitem{CommentLag} Due to the lag between the tip voltage and the peak intensity of the injection of seed CVRs caused by the intrinsic dynamics of the charged tangle, this procedure systematically overestimates $\tau$ by as much as $\sim 0.2$\,s; this translates into an overestimation of the initial radius of the seed CVRs $\bar{R_0} \approx 0.5$\,$\mu$m by as much as a factor of two. Yet, these uncertainties are insignificant for experiments with high $U>100$\,V (so that $R(d)\sim 5$\,$\mu$m $\gg R_0$), and do not affect our main observations of the changes in the shape and amplitude of the collector peaks.
\bibitem{WalmsleyJLTPPobell} P. M. Walmsley, A. A. Levchenko, S. E. May, and A. I. Golov, J. Low Temp. Phys. {\bf 146}, 511 (2007).
\bibitem{pure4He} Helium gas with \3 concentration less than $2\times 10^{-11}$ was produced in Lancaster University by the technique published in P. C. Hendry and P. V. E. McClintock, Cryogenics {\bf 27}, 131 (1987).
\bibitem{CommentBuoyantRings} This is similar to the dynamics of classical buoyant vortex rings \cite{Turner1957,buoyant2010}.
\bibitem{Turner1957} J. S. Turner, Proc. Royal Soc. London. Series A, {\bf 239}, 61-75 (1957).
\bibitem{buoyant2010} D. Bond, H. Johari, Exp. Fluids {\bf 48}, 737 (2010).
\bibitem{GolovIshimoto} A. Golov and H. Ishimoto, J. Low Temp. Phys. {\bf 113}, 957 (1998).
\bibitem{Supplement} Supplementary materials to this Letter.
\bibitem{CommentPile-ups} These secondary collisions were envisaged in \cite{Bradley2005,Fujiyama2010}.
\bibitem{Fujiyama2010} S. Fujiyama {\it et al.}, Phys. Rev. B {\bf 81}, 180512(R) (2010).
\bibitem{1-d} In our experiments, CVRs' motion very near the injector is not quite one-dimensional, but the number of reconnections is small there, hence the end result should not be strongly affected by this approximation.
\bibitem{Paul_sigma} Our preliminary calculations of the interaction between two parallel vortex rings of slightly different radii
%(especially effective for the case of unidirectional rings with small relative velocity $\sim \delv$, as two rings spend long time at a close distance)
show that reconnections occur with the cross-section $\sigma _1 \approx 0.6 R^2$, while a noticible perturbation of the directions of motion and velocities (even without reconnections) occurs with the cross-section even larger than $\sigma _1 = 4\pi R^2$.
\bibitem{sigma2} We assume that, upon a collision of two primary CVRs of radii $R$ each carrying impulse $\rho\kappa\pi R^2$, they merge into a circular loop with impulse $2\rho\kappa\pi R^2$ and hence radius $\sqrt{2}R$.
\end{thebibliography}
\end{document}