The area rule for circulation in three-dimensional turbulence

TL;DR

This paper investigates the Area Rule in three-dimensional turbulence, analyzing how circulation PDF depends on minimal loop area through high-resolution simulations, revealing subtle shape effects and a connection to minimal surfaces.

Contribution

It provides the first detailed assessment of the Area Rule's robustness in 3D turbulence, including effects of loop shape and minimal surface considerations, using high-resolution DNS data.

Findings

Circulation moments for rectangular loops closely match square loops with minor differences.

Aspect ratio effects are similar to Gaussian Random Field results, with differences under 5%.

Circulation statistics around minimal surfaces in 3D match planar loops when normalized appropriately.

Abstract

An important idea underlying a plausible dynamical theory of circulation in three-dimensional turbulence is the so-called Area Rule, according to which the probability density function (PDF) of the circulation around closed loops depends only on the minimal area of the loop, not its shape. We assess the robustness of the Area Rule, for both planar and non-planar loops, using high-resolution data from Direct Numerical Simulations. For planar loops, the circulation moments for rectangular shapes match those for the square with only small differences, these differences being larger when the aspect ratio is further from unity, and when the moment-order increases. The differences do not exceed about $5\%$ for any condition examined here. The aspect-ratio dependence observed for the second-order moment are indistinguishable from results for the Gaussian Random Field (GRF) with the same two-point correlation function (for which the results are order-independent by construction). When normalized by the SD of the PDF, the aspect ratio dependence is even smaller $(< 2\%)$ but does not vanish unlike for the GRF. We obtain circulation statistics around minimal area loops in three dimensions and compare them to those of a planar loop circumscribing equivalent areas, and we find that circulation statistics match in the two cases only when normalized by an internal variable such as the SD. This work highlights the hitherto unknown connection between minimal surfaces and turbulence.