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/

Zendy Plus

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

Zendy Tools

Zendy Open

This work presents a formalism to improve the predictive accuracy of physicalmodels by learning generalizable augmentations from sparse data. Building onrecent advances in data-driven turbulence modeling, the present approach,referred to as Learning and Inference assisted by Feature-space Engineering(LIFE), is based on the hypothesis that robustness and generalizability demanda meticulously-designed feature space that is informed by the underlyingphysics, and a carefully constructed features-to-augmentation map. The criticalcomponents of this approach are: (1) Maintaining consistency across thelearning and prediction environments; (2) Tightly-coupled inference andlearning by constraining the augmentation to be learnable throughout theinference process; (3) Identification of relevant physics-informed features inappropriate functional forms to enable significant overlap in feature space fora wide variety of cases to promote generalizability; (4) Maintaining explicitcontrol over feature space to change the augmentation function behavior only inthe vicinity of available datapoints. To demonstrate the viability of thisapproach, it is used in the modeling of bypass transition. The augmentation isdeveloped on skin friction data from two flat plate cases from the ERCOFTACdataset. Piecewise linear interpolation on a structured grid in feature-spaceis used as a sample functional form for the augmentation to demonstrate thecapability of localized learning. The augmented model is then applied to avariety of flat plate cases which are characterized by different freestreamturbulence intensities, pressure gradients, and Reynolds numbers. Thepredictive capability of the augmented model is also tested on single-stagehigh-pressure-turbine cascade cases, and the model performance is analyzed fromthe perspective of information contained in the feature space.

Generalizable Physics-constrained Modeling using Learning and Inference assisted by Feature Space Engineering