CFC2023

Student

Digital twin for freeze drying

  • Kamenik, Blaz (University of Maribor, Faculty of Mechanical Engineering)
  • Hribersek, Matjaz (University of Maribor, Faculty of Mechanical Engineering)
  • Ravnik, Jure (University of Maribor, Faculty of Mechanical Engineering)
  • Zadravec, Matej (University of Maribor, Faculty of Mechanical Engineering)

Please login to view abstract download link

Freeze-drying is a form of drying in which water is removed from a frozen product by a low-pressure sublimation process. The product is usually found in pharmacies in vials. Experimentally determining the drying kinetics of a product is often time consuming and uses expensive ingredients. There also remains the problem of transferring drying cycles between devices of different sizes, since local hydrodynamic conditions may be different. To model the time-dependent drying process, various geometric approximations for vials are used, ranging from 0D models to 1D models to 2D axisymmetric vial models. In order to evaluate the local hydrodynamic conditions within the freeze dryer, computational fluid dynamics (CFD) is increasingly used to model the process. Examples of such an approach are the numerical models that consider the effects of the geometry and position of the valves, the deposition of ice on the cold surfaces of the condenser, and the choked flow. Currently, there is a lack of coupled models that would allow direct evaluation of the effects of local hydrodynamic conditions on the drying of the product in all vials in the drying chamber. In this work, we present a numerical model for the transport phenomena during freeze-drying that simulates the drying kinetics of a product in each vial. For this purpose, we use a specific 1D product model and a hydraulic model for stopper resistance coupled with the time-dependent 3D CFD solution of the flow field in the drying chamber. In this study, laboratory freeze-drying model with two shelves installed was used. As can be seen from the calculated local pressure distribution across the shelf, there are significant local pressure differences between the vials on the bottom shelf. The pressure is highest in the middle of the shelf and decreases towards the sides of the shelf. As can be seen, pressure differences of up to 5 Pa are observed, which can significantly affect the pressure difference between the sublimation front and the drying chamber under the targeted conditions of the low-pressure system, resulting in lower mass flow rates for the vials in the middle shelf positions. The computational results of the 1D model coupled to the 3D CFD model show that the model is capable of resolving variations in local chamber pressure conditions and local vial drying kinetics. The modelling approach provides a reliable digital twin for the freeze drying process.