Search citation statements
Paper Sections
Citation Types
Year Published
Publication Types
Relationship
Authors
Journals
The long lasting demand for better turbulence models and the still prohibitively computational cost of high‐fidelity fluid dynamics simulations, like direct numerical simulations and large eddy simulations, have led to a rising interest in coupling available high‐fidelity datasets and popular, yet limited, Reynolds averaged Navier–Stokes simulations through machine learning (ML) techniques. Many of the recent advances used the Reynolds stress tensor or, less frequently, the Reynolds force vector as the target for these corrections. In the present work, we considered an unexplored strategy, namely to use the modeled terms of the Reynolds stress transport equation as the target for the ML predictions, employing a neural network approach. After that, we solve the coupled set of governing equations to obtain the mean velocity field. We apply this strategy to solve the flow through a square duct. The obtained results consistently recover the secondary flow, which is not present in the baseline simulations that used the model. The results were compared with other approaches of the literature, showing a path that can be useful in the seek of more universal models in turbulence.
The long lasting demand for better turbulence models and the still prohibitively computational cost of high‐fidelity fluid dynamics simulations, like direct numerical simulations and large eddy simulations, have led to a rising interest in coupling available high‐fidelity datasets and popular, yet limited, Reynolds averaged Navier–Stokes simulations through machine learning (ML) techniques. Many of the recent advances used the Reynolds stress tensor or, less frequently, the Reynolds force vector as the target for these corrections. In the present work, we considered an unexplored strategy, namely to use the modeled terms of the Reynolds stress transport equation as the target for the ML predictions, employing a neural network approach. After that, we solve the coupled set of governing equations to obtain the mean velocity field. We apply this strategy to solve the flow through a square duct. The obtained results consistently recover the secondary flow, which is not present in the baseline simulations that used the model. The results were compared with other approaches of the literature, showing a path that can be useful in the seek of more universal models in turbulence.
Complex turbulent flows with large-scale instabilities and coherent structures pose challenges to both traditional and data-driven Reynolds-averaged Navier–Stokes methods. The difficulty arises due to the strong flow-dependence (the non-universality) of the unsteady coherent structures, which translates to poor generalizability of data-driven models. It is well-accepted that the dynamically active coherent structures reside in the larger scales, while the smaller scales of turbulence exhibit more “universal” (generalizable) characteristics. In such flows, it is prudent to separate the treatment of the flow-dependent aspects from the universal features of the turbulence field. Scale resolving simulations (SRS), such as the partially averaged Navier–Stokes (PANS) method, seek to resolve the flow-dependent coherent scales of motion and model only the universal stochastic features. Such an approach requires the development of scale-sensitive turbulence closures that not only allow for generalizability but also exhibit appropriate dependence on the cut-off length scale. The objectives of this work are to (i) establish the physical characteristics of cut-off dependent closures in stochastic turbulence; (ii) develop a procedure for subfilter stress neural network development at different cut-offs using high-fidelity data; and (iii) examine the optimal approach for the incorporation of the unsteady features in the network for consistent a posteriori use. The scale-dependent closure physics analysis is performed in the context of the PANS approach, but the technique can be extended to other SRS methods. The benchmark “flow past periodic hills” case is considered for proof of concept. The appropriate self-similarity parameters for incorporating unsteady features are identified. The study demonstrates that when the subfilter data are suitably normalized, the machine learning based SRS model is indeed insensitive to the cut-off scale.
Turbulence closure modeling using machine learning is at an early crossroads. The extraordinary success of machine learning (ML) in a variety of challenging fields had given rise to an expectation of similar transformative advances in the area of turbulence closure modeling. However, by most accounts, the current rate of progress toward accurate and predictive ML-RANS (Reynolds Averaged Navier-Stokes) closure models has been very slow. Upon retrospection, the absence of rapid transformative progress can be attributed to two factors: the
underestimation of the intricacies of turbulence modeling and the overestimation of ML’s ability to capture all features without employing targeted strategies. To pave the way for more
meaningful ML closures tailored to address the nuances of turbulence, this article seeks to
review the foundational flow physics to assess the challenges in the context of data-driven
approaches. Revisiting analogies with statistical mechanics and stochastic systems, the key
physical complexities and mathematical limitations are explicated. It is noted that the current
ML approaches do not systematically address the inherent limitations of a statistical approach
or the inadequacies of the mathematical forms of closure expressions. The study underscores
the drawbacks of supervised learning-based closures and stresses the importance of a more
discerning ML modeling framework. As ML methods evolve (which is happening at a rapid
pace) and our understanding of the turbulence phenomenon improves, the inferences expressed
here should be suitably modified.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.