Thumbnail Image

Design of an Axial Flow Bioreactor for Tissue Regeneration

Raja Bhaskar, Prasana
Modeling of bioreactors using Computational Fluid Dynamics (CFD) tools have been adapted for several bioreactor configurations such as flow-through, parallel-flow, and rotary. In this study, modeling was mainly performed to analyze the hydrodynamic characteristics, such as shear stress, and pressure drop, and nutrient distribution. The axial flow bioreactor configuration offer several advantages, such as convection-driven nutrient distribution and ability to operate at high flow rates. Hence, this configuration was selected for simulation studies. The geometry of the bioreactor was optimized using COMSOL 4.2 to ensure uniform shear stress and nutrient distribution throughout scaffold. The bioreactor was fabricated in-house for experimental studies. Experimental validation of simulation results were done by measuring pressure drop across the bioreactor and analyzing residence time distribution (RTD) in the bioreactor. The hold-up volume of the bioreactor (scaled up) was minimized by increasing the semi-angle of the cone. However, increasing the semi-angle reduced the distribution of nutrients in the outer regions of the scaffold. Uniform nutrient distribution was achieved by incorporating a distribution system at the entrance of the bioreactor. The simulation results showed increase in pressure drop across the bioreactor with increased fluid flow rate and decreased scaffold pore size. Changing the inlet or outlet diameter of the bioreactor had little to no effect on hydrodynamic characteristics or nutrient distribution. The bioreactor configuration with minimum hold-up volume, uniform nutrient and shear stress distribution was selected for experimentation. Experimental measurement of pressure drop across the bioreactor without the scaffold showed good agreement with the simulation results. However, experimental pressure drop across the bioreactor with scaffold showed deviation from simulation results. This was attributed to the skinny layer on top of the scaffold. The residence time distribution experiments suggested a decrease in the dead volume of the bioreactor with the addition of the distribution system. Comparison of nutrient distribution in simulations with experimental RTD showed possible dead zones. Future studies should focus on modifying the distributor system to minimize dead volumes. Cell culture experiments must be conducted on the bioreactor to evaluate the effectiveness of the bioreactor configuration.