Date on Master's Thesis/Doctoral Dissertation

7-2015

Document Type

Master's Thesis

Degree Name

M. Eng.

Department

Chemical Engineering

Committee Chair

Berson, Robert E. (Eric)

Committee Co-Chair (if applicable)

Willing, Gerold

Committee Member

O'Toole, Martin

Author's Keywords

Bioreactor; Computational fluid dynamics; Bone tissue engineering

Subject

Bioreactors; Computational fluid dynamics; Bone regeneration; Tissue engineering

Abstract

Through recent years, many novel bioreactor designs for tissue engineering have emerged. The Tubular Perfusion System (TPS) is one such bioreactor for bone and cartilage tissue engineering. The TPS is a perfusion type bioreactor, in which media flows through a bed of scaffold beads. The advantage of this system over similar designs is that it requires no special fabrication techniques and operates at low pressure. Experimental trials have shown increased osteoblastic differentiation and mineralization of bone tissue grown using the TPS versus static culture. The effectiveness of the TPS in promoting osteoblastic differentiation is believed to be due to enhanced nutrient transport and exposure of the scaffold particles to shear stress. The purpose of this thesis was to quantify these two factors using computational fluid dynamics and to study the effects of altering the packing regime of the bed, the media flow rate, and the scaffold materials. The study successfully identified an average shear exposure of 0.13-0.26 dyn/cm2 for all cases under normal operating conditions. The highest average shear stress recorded corresponds to a chamber to bead diameter ratio (D/d) of 2.821. This case is unique because the geometry contains interior spheres. An analysis of flow within the scaffold particles determined the relative contributions of diffusive and advective transport within the scaffolds, with average Peclet numbers ranging from 0.05 to 0.30. In more permeable tri-calcium phosphate scaffolds, advective transport plays a much more significant role. Both phenomena were found to be highly geometry dependent, with different distributions for each packing regime simulated

Download

Share

COinS