Simulation and experiments of low pressure water vapor flows applied to freeze-drying

Arnab Ganguly, Purdue University

Abstract

The design of freeze-drying equipment used today is based largely on empirical knowledge. The freeze-dryers used in product manufacturing have changed little since the mid-20th century. A typical pharmaceutical freeze-drying cycle is run as a batch process and is thus, both, time and energy intensive. The equipment design cycle adopted today relies heavily on post-concept tests. This limits the design changes that can be adopted as an outcome of these tests. For example, a single 6 hour drying test run on a production scale dryer may cost up to $1000. Moreover, the associated costs are passed onto the customers. The current work is an attempt at setting up a comprehensive modeling framework that can predict the performance of a pharmaceutical freeze-dryer in the pre-concept design stage. In order to develop a modeling framework, a low temperature water vapor molecular model that could be used consistently for the temperature range used in a freeze-drying process was first developed. The variable hard sphere (VHS) model was found to predict the temperature dependence of viscosity well. Suitable values of the VHS parameters were determined. At a reference temperature of 273K, the effective diameter of the water molecule was found to be 5.78 A˙. The value of the viscosity-temperature exponent equal to 1.0 allows reproducing the viscosity coefficient variation up to a temperature of 373 K within 0.5% and up to 423 K within 3% of the internationally accepted standard. A typical freeze-dryer may be sub-divided into three sections based on functionality, (a) the product chamber, (b) the duct, and (c) the condenser chamber. The developed model was then applied to modeling the flow structure in each of these three sections. Experimental measurements using tunable diode laser absorption spectroscopy revealed a significant deviation of 22.6% between the measured mass flow rate data and gravimetrically computed water vapor mass flow rates. It was found that the presence of clean in the place (CIP) /sterilize in place (SIP) piping and the small length to diameter ratio of the duct led to the deviation from the assumed analytical velocity profile in the duct. The demand for a paradigm shift towards continuous freeze-drying in pre-filled syringes has engendered the need to understand the heat transfer mechanisms in the absence of contact between the vial and the shelf. Heat and mass transfer processes in suspended vials has been studied experimentally to understand the role played by the different modes of heat transfer in a laboratory scale freeze dryer. The role played by the shelf and the dryer door in the heat exchange with the product are discussed. Experimental measurements of heat flux in a laboratory scale free-dryer reveal that the convective component of heat transfer in the product chamber cannot be ignored at higher pressures that are more commonly used in primary drying, something that has traditionally been assumed for suspended vial heat transfer analysis. Improving condenser designs used today is essential for achieving reduced cycle times. Analysis of vapor and ice dynamics in freeze-dryer condensers are discussed with an eye towards understanding the flow structure in a condenser chamber. Direct Simulation Monte Carlo (DSMC) techniques are applied to model the physical processes that accompany low pressure vapor flow in the condenser chamber. Relevant metrics for comparing condenser performance are defined. These metrics were used to compare the performance of two different condenser designs. The steady state DSMC solution was then used to predict the ice formation on the condenser coils. The DSMC simulations demonstrate that by tailoring the condenser topology to the flow-field structure of the water vapor jet expanding into a low-pressure reservoir, it is possible to significantly increase water vapor removal rates and improve the overall efficiency of freeze-drying processes.

Degree

M.S.A.A.

Advisors

Alexeenko, Purdue University.

Subject Area

Aerospace engineering|Mechanical engineering|Pharmacy sciences

COinS