Prediction of thermal conductivity of liquids by application of mixing laws and symbolic regression

Thermal conductivity is a key property for the sizing of many processes equipments, such as exchangers, which allow to transfer thermal energy from one fluid to another, without mixing them. In recent years, we have worked on the application of machine learning (ML) tools on data extracting from existing databases and/or from the literature, and we have recently proposed models to predict the thermal conductivity of hydrocarbons and oxygenates in the liquid phase as a function of temperature.1

The objective of the proposed internship is to extend the previous approach to the case of mixtures. Thus, the work will begin with the establishment of a thermal conductivity database for mixtures, data that will serve as support for the development of predictive numerical approaches based on ML. Thermal conductivity data will be extracted from both existing databases and the literature, these data will be merged, curated and formatted. Two prediction approaches will be considered for this work:

• The first will consist of using models recently developed to predict thermal conductivity of hydrocarbons and oxygenates,1 to feed mixing laws available in the literature.

• The second approach will focus on the use of symbolic regression to identify, by regression of reference data, mathematical expressions allowing the best modeling of this data.

The intern will be supervised by engineers from the “Thermodynamics and Molecular Modeling” department of IFPEN and will also have the opportunity to work with teams from the University of Thessaly in Greece.

Internship level: M2

Internship duration: 6 months

Period: first semester 2025

Discipline: Chemoinformatics

Skills: Python programming, Chemistry, Fluent in english

Administrative information

Location: IFP Energies nouvelles

Address: 1-4 avenue de Bois Préau - 92500 Rueil-Malmaison

Division: Applied Physico-Chemistry and Mechanics

Internship supervisers: Benoît Creton, Carlos Nieto, Véronique Lachet

email of contact : benoit.creton@ifpen.fr -> Please send us a CV, cover letter.