Computational fluid dynamics analysis of freeze drying process and equipment

Nikhil P Varma, Purdue University

Abstract

Freeze drying is an important, but expensive, inefficient and time consuming process in the pharmaceutical, chemical and food processing industries. Computational techniques could be a very effective tool in predictive design and analysis of both freeze drying process and equipment. This work is an attempt at using Computational Fluid Dynamics(CFD) and numerical simulations as a tool for freeze drying process and equipment design. Pressure control is critical in freeze dryers, keeping in view the product stability. In industrial freeze dryers, loss of pressure control can lead to loss of an entire batch. Pressure variation within the chamber could also lead to batch inhomogeneity, especially in industrial scale dryers. The low-pressure environment and the relatively small flow velocities make it difficult to quantify the flow structure experimentally. The current work presents a three-dimensional multi-species computational fluid dynamics model for vapor flow in a laboratory scale freeze-dryer validated with experimental data and analytical expressions. The model accounts for the presence of a non-condensable gas such as nitrogen or air using a continuum multi-species model. The flow structure at different sublimation rates, chamber pressures and shelf-gaps are systematically investigated. Emphasis has been placed on accurately predicting the pressure variation across the subliming front. It was found that while the pressure variation increased linearly with sublimation rate in the range of 0.5 kg/hr/m2 to 1.3 kg/hr/m2, the variation was more sensitive at shelf gaps approaching about at 2.1 cm and negligible at gaps close to 9 cm. While the results are found to agree within 10\% of measurements made for the range of shelf gaps and sublimation rates investigated here, the analytical solution is found to be more accurate for smaller shelf gaps. The current work presents an important validation case motivating broader use of CFD in optimizing process and equipment design. A critical component of the freeze drying system is the chamber to pressure duct. A well designed duct, apart from providing smooth flow and lower chamber pressures, should be able to accommodate the peak flow rates without the possibility choking. Here we use computational fluid dynamics as a tool to predict the minimum controllable chamber pressure, maximum sublimation rate, the onset of choking and to suggest better freeze dryer geometry. The main findings include an improved performance concept (IPC) that allowed 11\% improvement in the sublimation rate. The IPC offered a lower minimum controllable pressure at all sublimation rates. A maximum projected improvement because of the combined effect of lower controllable pressure and improved flow throughput is estimated at 14.7\% for the IPC. For the lower condenser temperature, the maximum performance improvement of 8\% was observed at the 50mTorr chamber pressure. Heat transfer to the product is one of the most inefficient step in the Freeze Drying process. While a higher product temperature ensures faster sublimation, it is necessary to maintain it below the collapse temperature for product stability. A quick and reasonable accurate model could accelerate the process design cycle or at the least reduce the number of experimental runs required. Here we present a simplified unsteady heat and mass-transfer model which can be adapted for different equipments and product geometry. The model is compared against experiments, the effect of slice geometry is discussed and the application in a continuous freeze dryer is demonstrated. The model was found to accurately predict the product temperature. However the sublimation rate predictions deviated considerably form experiential studies. It is speculated that the porous structure of the dried cake could play a significant role in the sublimation rates and hence a more complicated model might be required for accurate predictions.

Degree

M.S.A.A.

Advisors

Alexeenko, Purdue University.

Subject Area

Chemical engineering

COinS