/ theme
Aerodynamic surrogate modelling
Latent-space and neural-operator surrogates for high-dimensional pressure and aerodynamic fields, from transonic wings to propellers.
Read moreUC3M · Aerospace Engineering
Rodrigo
Castellanos
Assistant Professor · Department of Aerospace Engineering · Universidad Carlos III de Madrid
Data-efficient AI for aerodynamic modelling, flow control, and aerospace design.
I develop machine-learning and data-driven methods for fluid mechanics and aerodynamics, with a focus on surrogate modelling, multifidelity learning, active flow control, aerodynamic optimization, and physics-aware AI. My work combines high-fidelity simulations, wind-tunnel experiments, reduced-order models, and modern scientific machine learning to make aerodynamic prediction more efficient and reliable.
- PositionAssistant Professor, UC3M
- PhDFluid Mechanics · Cum Laude
- RecognitionOutstanding Thesis Award
- Publications12+ journal articles
- Impact (Scholar)Loading…
- StaysTU Delft · Centrale Nantes
/ 01 — Research at a glance
Four lines of work, one principle: learn aerodynamics from scarce, expensive data.
/ theme
Multifidelity & data-efficient learning
Combining scarce high-fidelity data with cheaper low-fidelity sources through transfer learning, Gaussian processes, and active sampling.
Read more/ theme
Flow control & optimization
Genetic, model-based and ML-driven strategies for active flow control, drag reduction, and convective heat-transfer enhancement.
Read more/ theme
Experimental fluid mechanics
Synchronized heat-flux and velocity measurements, infrared thermography, PIV/PTV, and wall-bounded turbulent flows.
Read more/ Positioning
“High-fidelity aerodynamic data are expensive. My research develops data-efficient, physics-aware machine-learning methods to build reliable surrogate models from scarce simulations and experiments.”
/ Collaborate
Looking for collaborations or supervision?
I welcome motivated students and academic or industrial partners interested in AI-driven aerodynamics, surrogate modelling, and flow control.