Developing hybrid Equations of State

Models for predicting thermophysical properties are central to the design and optimization of chemical engineering processes. One interesting class of such models are Equations of State (EoS), which can describe essentially all thermophysical properties over a wide temperature and pressure range, including supercritical states. Alongside established physical EoS, data-driven methods from machine learning (ML) now open up exciting new possibilities. Particularly noteworthy are hybrid models that combine physics-based EoS with ML methods. In my project, I therefore develop, train, and test hybrid equations of state for predicting thermophysical properties of pure components and mixtures.

Teaching

• Fundamentals of Machine Learning for Engineers (SuSe 2026)
• Modelling, Simulation and Optimization in Process Engineering (WS 2026/27)
   

Office Hours

By appointment

Publications

• F. Fleckenstein, M. Wagner, T. Bernet, E.A. Müller, S. Stephan: Examination of inconsistencies in the physical modeling of vapor–liquid interfaces of strongly non-ideal mixtures, Physical Chemistry Chemical Physics, 27 (2025) 23966-23985. [doi]
   

Presentations

-

Posters

-