Luca Muscara'

Ph.D. candidate in Ingegneria Aerospaziale , 37th cycle (2021-2024)
Department of Mechanical and Aerospace Engineering (DIMEAS)



Research topic

Machine learning for multiphysics problems


Research interests

Computational Fluid Dynamics
Turbulent Flows


Having earned a Master's degree in Aerospace Engineering on April 6, 2020, he focused his academic journey on propulsion systems, particularly Hybrid Rocket Engines (HREs). His master's thesis involved developing a numerical model to analyze combustion instability in HREs, a critical area lacking a comprehensive theory. This endeavor required integrating three physical models—quasi-1D gas dynamics, chemical, and thermal models—due to the intricate multi-physics nature of the field.

Following graduation, he pursued scientific advancement through an eight-month research scholarship, 'Sviluppo di modelli numerici per endoreattori ibridi,' starting from September 7, 2020, at Politecnico di Torino . His project, supported by ASI with Politecnico di Torino and AVIO Spa collaboration, focuses on developing a numerical model for HREs using liquefying propellants. This endeavor demands improvements to the thermal model to accommodate the presence of the melt layer at the combustion surface.
Throughout this journey, he honed his skills in high-resolution spatial discretization methods, time integration schemes, Finite Volume Methods for dynamic mesh, and numerical heat transfer. A significant challenge in his research involved optimizing computational efficiency, particularly in the chemical sub-model. To address this, he explored the integration of Machine Learning (ML) techniques, employing Artificial Neural Networks (ANNs) as a surrogate chemistry model to significantly accelerate simulations while maintaining accuracy.

In November 2021, he embarked on a new chapter in his academic journey by initiating a Ph.D. program with a specific focus on Machine Learning for multiphysics problems, exploring potential applications in fluid mechanics and turbulent flow. Currently, he is actively employing the field inversion technique, advancing the integration of ML with turbulence models to enhance the predictive capabilities of Reynolds-Averaged Navier–Stokes (RANS) models, enabling accurate simulations with reduced computational resources.

Research topics

  • Machine learning for multiphysics problems



Goal 4: Quality education
Goal 8: Decent work and economic growth
Goal 9: Industry, Innovation, and Infrastructure



Master of Science

  • Endoreattori. A.A. 2022/23, INGEGNERIA AEROSPAZIALE. Collaboratore del corso
MostraNascondi A.A. passati


Research groups


Latest publications View all publications in Porto@Iris