About us
Espace utilisateur
INSTN offers more than 40 diplomas from operator level to post-graduate degree level. 30% of our students are international students.
Professionnal development
Professionnal development
Find a training course
INSTN delivers off-the-self or tailor-made training courses to support the operational excellence of your talents.
Human capital solutions
At INSTN, we are committed to providing our partners with the best human capital solutions to develop and deliver safe & sustainable projects.
Home   /   Thesis   /   Contribution of learning methods to separation and recycling processes: towards a numrical twin of a mixer-settler

Contribution of learning methods to separation and recycling processes: towards a numrical twin of a mixer-settler

Engineering sciences Mathematics - Numerical analysis - Simulation Mechanics, energetics, process engineering


Low-carbon energies (lithium batteries, photovoltaics, wind power) are largely based on rare earths (Dy, Nd, Pr, ...) and metals (Co, Ni, ...). Numerous studies are currently devoted to the development of liquid-liquid extraction processes adapted to their recycling, but their industrial transposition remains a major technological obstacle, which the thesis intends in part to overcome.
The performance of solvent extraction processes strongly depends on the exchange surface available between the two phases (generally an organic solvent and the aqueous dissolution medium) and through which the actual extraction takes place. However, this available surface is difficult to measure and predict because of the complexity of the phenomena involved. It is in fact strongly correlated with the physicochemical properties of the liquid-liquid system considered and with the properties of the turbulent flow generated by the stirring system used to create an interface between the liquids (dipersion of a phase in the form of droplets).
The objective of the thesis is to explore the potential of machine learning methods for the prediction of the exchange surface in a stirred tank, representative of a mixer-settler.
The experimental study will focus on the determination of the drop size distribution in the stirred tank (optical methods and image analysis), the numerical part will include the CFD simulation of the turbulent flow and the development of the learning model.
The thesis is located in Marcoule CEA research center, near Avignon, in a multidisciplinary team with a strong focus on processes for green transition. The applicant we are looking for is an engineer and/or master 2 having a generalist profile and interested in playing an active role in this field. At the end of this Ph.D., the candidate will have a first experience in machine learning methods applied to the present issues of recycling in circular economy. Such a type of versatile profile will undoubtedly be an asset for a future career in academia as well as in industry.


Département de recherche sur les procédés pour la mine et le recyclage du combustible
Service des Technologies Durables pour le Cycle des matières
Laboratoire de développement de procédés pour le Recyclage et la Valorisation
Top envelopegraduation-hatlicensebookuserusersmap-markercalendar-fullbubblecrossmenuarrow-down