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

Experimental measurements and numerical simulations of turbulent flows arecharacterised by a trade-off between accuracy and resolution. In this study, webridge this gap using Physics Informed Neural Networks (PINNs) constrained bythe Reynolds-Averaged Navier-Stokes (RANS) equations and accurate sparsepointwise mean velocity measurements for data assimilation (DA). Firstly, byconstraining the PINN with sparse data and the under-determined RANS equationswithout closure, we show that the mean flow is reconstructed to a higheraccuracy than a RANS solver using the Spalart-Allmaras (SA) turbulence model.Secondly, we propose the SA turbulence model augmented PINN (PINN-DA-SA), whichoutperforms the former approach - up to 73% reduction in mean velocityreconstruction error with coarse measurements. The additional SA physicsconstraints improve flow reconstructions in regions with high velocity andpressure gradients and separation. Thirdly, we compare the PINN-DA-SA approachto a variational data assimilation using the same sparse velocity measurementsand physics constraints. The PINN-DA-SA achieves lower reconstruction erroracross a range of data resolutions. This is attributed to discretisation errorsin the variational methodology that are avoided by PINNs. We demonstrate themethod using high fidelity measurements from direct numerical simulation of theturbulent periodic hill at Re=5600.

Turbulence model augmented physics informed neural networks for mean flow reconstruction