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

Over the past few years, deep learning methods have proved to be of greatinterest for the computational fluid dynamics community, especially when usedas surrogate models, either for flow reconstruction, turbulence modeling, orfor the prediction of aerodynamic coefficients. Overall, exceptional levels ofaccuracy have been obtained, but the robustness and reliability of the proposedapproaches remain to be explored, particularly outside of the confidence regiondefined by the training dataset. In this contribution, we present a twinauto-encoder network for incompressible laminar flow reconstruction withuncertainty estimation around 2D obstacles. The proposed architecture istrained over a dataset composed of numerically-computed laminar flows around12,000 random shapes, and naturally enforces a quasi-linear relation between ageometric reconstruction branch and the flow prediction decoder. Based on thisfeature, two uncertainty estimation processes are proposed, allowing either abinary decision (accept or reject prediction), or proposing a confidenceinterval along with the flow quantities prediction (u,v,p). Results overdataset samples as well as unseen shapes show a strong positive correlation ofthis reconstruction score to the mean-squared error of the flow prediction.Such approaches offer the possibility to warn the user of trained models whenthe provided input shows too large deviation from the training data, making theproduced surrogate model conservative for fast and reliable flow prediction.

A twin-decoder structure for incompressible laminar flow reconstruction with uncertainty estimation around 2D obstacles