Tipo di tesi
Tesi di laurea magistrale
Titolo
High fidelity data exploitation for RANS turbulence modelling enhancements: machine learning symbolic regression for localized corrections
Dipartimento
INGEGNERIA CIVILE E INDUSTRIALE
Corso di studi
INGEGNERIA AEROSPAZIALE
Parole chiave
- cfd
- computational fluid dynamics
- symbolic regression
Data inizio appello
27/05/2026
Riassunto (Inglese)
In Computational Fluid Dynamics, a trade-off exists between computational resources and solution accuracy. In recent years, many data-driven approaches have emerged to bridge this gap. Building upon the work of Schmelzer et al. (2019), this thesis enhances the accuracy of RANS simulations by including symbolic corrections to the Boussinesq hypothesis. Unlike the original global approach, these corrections are applied locally only to specific regions of the domain via a filtering process. By reducing the amount of data to be processed, this method decreases the time required to identify the symbolic model and leads to more accurate results.
