English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Two models based on convolutional neural networks are trained to predict thetwo-dimensional velocity-fluctuation fields at different wall-normal locationsin a turbulent open channel flow, using the wall-shear-stress components andthe wall pressure as inputs. The first model is a fully-convolutional neuralnetwork (FCN) which directly predicts the fluctuations, while the second onereconstructs the flow fields using a linear combination of orthonormal basisfunctions, obtained through proper orthogonal decomposition (POD), hence namedFCN-POD. Both models are trained using data from two direct numericalsimulations (DNS) at friction Reynolds numbers $Re_{\tau} = 180$ and $550$.Thanks to their ability to predict the nonlinear interactions in the flow, bothmodels show a better prediction performance than the extended proper orthogonaldecomposition (EPOD), which establishes a linear relation between input andoutput fields. The performance of the various models is compared based onpredictions of the instantaneous fluctuation fields, turbulence statistics andpower-spectral densities. The FCN exhibits the best predictions closer to thewall, whereas the FCN-POD model provides better predictions at largerwall-normal distances. We also assessed the feasibility of performing transferlearning for the FCN model, using the weights from $Re_{\tau}=180$ toinitialize those of the $Re_{\tau}=550$ case. Our results indicate that it ispossible to obtain a performance similar to that of the reference model up to$y^{+}=50$, with $50\%$ and $25\%$ of the original training data. Thesenon-intrusive sensing models will play an important role in applicationsrelated to closed-loop control of wall-bounded turbulence.

Convolutional-network models to predict wall-bounded turbulence from wall quantities