Source author record

Diego A. Donzis

Diego A. Donzis appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

physics.flu-dyn nlin.CD physics.comp-ph

Catalog footprint

What is connected

3works

3topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

Direct Numerical Simulations of turbulent flows using high-order Asynchrony-Tolerant schemes: accuracy and performance

Direct numerical simulations (DNS) are an indispensable tool for understanding the fundamental physics of turbulent flows. Because of their steep increase in computational cost with Reynolds number ($R_λ$), well-resolved DNS are realizable only on massively parallel supercomputers, even at moderate $R_λ$. However, at extreme scales, the communications and synchronizations between processing elements (PEs) involved in current approaches become exceedingly expensive and are expected to be a major bottleneck to scalability. In order to overcome this challenge, we developed algorithms using the so-called Asynchrony-Tolerant (AT) schemes that relax communication and synchronization constraints at a mathematical level, to perform DNS of decaying and solenoidally forced compressible turbulence. Asynchrony is introduced using two approaches, one that avoids synchronizations and the other that avoids communications. These result in periodic and random delays, respectively, at PE boundaries. We show that both asynchronous algorithms accurately resolve the large-scale and small-scale motions of turbulence, including instantaneous and intermittent fields. We also show that in asynchronous simulations the communication time is a relatively smaller fraction of the total computation time, especially at large processor count, compared to standard synchronous simulations. As a consequence, we observe improved parallel scalability up to $262144$ processors for both asynchronous algorithms.

preprint2019arXiv

Universality and scaling in compressible turbulence

Turbulent flows, ubiquitous in nature and engineering, comprise fluctuations over a wide range of spatial and temporal scales. While flows with fluctuations in thermodynamic variables are much more common, much less is known about these flows than their incompressible counterparts in which thermodynamics is decoupled from hydrodynamics. A critical element in the study of the latter has been the concept of universal scaling laws which provides fundamental as well as practical information about the spatio-temporal behavior of these complex systems. Part of this success is due to the dependence on a single non-dimensional parameter, that is the Reynolds number. Universality in compressible flows, on the other hand, have proven to be more elusive as no unifying set of parameters were found to yield universal scaling laws. This severely limits our understanding of these flows and the successful development of theoretically sound models. Using results in specific asymptotic limits of the governing equations we show that universal scaling is indeed observed when the set of governing parameters is expanded to include internally generated dilatational scales which are the result of the driving mechanisms that produce the turbulence. The analysis demonstrate why previous scaling laws fail in the general case, and opens up new venues to identify physical processes of interest and create turbulence models needed for simulations of turbulent flows at realistic conditions. We support our results with a new massive database of highly-resolved direct numerical simulations along with all data available in the literature. In search of universal features, we suggest classes which bundle the evolution of flows in the new parameter space. An ultimate asymptotic regime predicted independently by renormalization group theories and statistical mechanics is also assssed with available data.

preprint2014arXiv

Regimes of nonlinear depletion and regularity in the 3D Navier-Stokes equations

The periodic $3D$ Navier-Stokes equations are analyzed in terms of dimensionless, scaled, $L^{2m}$-norms of vorticity $D_{m}$ ($1 \leq m < \infty$). The first in this hierarchy, $D_{1}$, is the global enstrophy. Three regimes naturally occur in the $D_{1}-D_{m}$ plane. Solutions in the first regime, which lie between two concave curves, are shown to be regular, owing to strong nonlinear depletion. Moreover, numerical experiments have suggested, so far, that all dynamics lie in this heavily depleted regime \cite{DGGKPV13}\,; new numerical evidence for this is presented. Estimates for the dimension of a global attractor and a corresponding inertial range are given for this regime. However, two more regimes can theoretically exist. In the second, which lies between the upper concave curve and a line, the depletion is insufficient to regularize solutions, so no more than Leray's weak solutions exist. In the third, which lies above this line, solutions are regular, but correspond to extreme initial conditions. The paper ends with a discussion on the possibility of transition between these regimes.