The Mechanical Engineering and Materials Department of the School of Engineering at the University of Navarra (Tecnun) is looking for a PhD candidate to carry out a doctoral thesis focused on the development of reduced models that are based on neural networks trained with results from Computational Fluid Dynamics (CFD) simulations. On one hand, the aim is to use convolutional neural networks (CNNs) that can provide results very close to those offered by CFD models but in a much shorter time, so that they can be used in dynamic simulations of systems and, ultimately, be incorporated into a Digital Twin. On the other hand, the possibilities that physics-informed neural networks (PINNs) can offer in this context are to be explored.
It is planned to develop reduced models for various water treatment processes, characterized by complex multiphase flows that may include biological and/or chemical reactions. However, the aim is that the methodologies developed for model reduction can be applied in any other field of engineering where the flow of one or several fluids plays a predominant role.
During the completion of the thesis the candidate will receive and acquire training in:
* Multiphase flow simulation.
* Turbulence modeling.
* Use of different types of neural networks for solving engineering problems.
Training activities for scientific, professional and personal development.
Working schedule:
* 7.75 hours per day.
* Flexible timetable: start between 8:00 and 9:30. Fridays with the possibility of a continuous working day.
* Summer timetable: Only mornings starting June 15th and finishing August 31st (5.5 hours).
Holidays:
* 23 working days + Christmas holidays (24-Dec to 2-Jan).
Degree: Master's Degree in: Industrial Engineering, Mechanical Engineering or Chemical Engineering.
Languages: English
IT knowledge:
* Computational Fluid Dynamics (CFD) simulation tools.
* Python.
* TensorFlow, Pytorch, Keras.
Others:
* Solid knowledge of Fluid Mechanics.
* Knowledge in Machine Learning and Deep Learning tools.
* Ability to work in a multidisciplinary team
Starting date: Early 2025