Multipoint Statistical Turbulent Dynamics from Hopf Equation Closures

Abstract

Obtaining accurate multipoint statistics of turbulence is computationally very expensive and therefore these statistics have remained largely unexplored from a theoretical standpoint. In this paper, (i) a first-principles-based closure of the $n$th-order structure function governing equation proposed by Sreenivasan & Yakhot (2021) is generalized to a closure of the velocity increment Hopf equation itself. Then (ii) the closure is further generalized to the $N$-point Hopf equation. Finally, (iii) an example of the method is provided to analytically determine the $3$-point structure function transition between the known $2$-point structure function and the $3$-point fusion rules from the closed $(N=3)$-point velocity increment Hopf equation. The analytical solution takes the form of a Batchelor interpolation and shows promising agreement with preliminary DNS data for the cases examined. Since the $N$-point velocity increment Hopf equation is closed, its solution can be numerically approximated. It is expected that similar methods, applied here to obtain the $2$-point structure function and $3$-point structure function transition, can be used to obtain further analytical predictions of various multipoint quantities to deepen our understanding of turbulence.