Available at: https://digitalcommons.calpoly.edu/theses/3431
Date of Award
8-2026
Degree Name
MS in Biomedical Engineering
Department/Program
Biomedical Engineering
College
College of Engineering
Advisor
Kristen Cardinal
Advisor Department
Biomedical Engineering
Advisor College
College of Engineering
Abstract
Dr. Kristen Cardinal’s Tissue Engineering Laboratory at California Polytechnic State University, San Luis Obispo develops tissue-engineered blood vessel mimics (BVMs) for preclinical testing of neurovascular devices. These BVMs are reliant on silicone vascular models that define the vessel geometry, support endothelial cell attachment, and allow devices to be evaluated in a controlled in vitro environment. Previous work in the lab established injection molding as a viable method for fabricating silicone vascular models and utilized acrylonitrile butadiene styrene (ABS) for sacrificial mandrels. However, this workflow relied on the use of acetone for mandrel dissolution, required single-use mandrels, had extended post-processing times, and was affected by changes in material availability and quality. The goal of this thesis was to develop and evaluate alternative sacrificial mandrel materials that could improve upon the current state of the in-house injection molding workflow.
The first aim of this thesis explored polyvinyl alcohol (PVA) as an alternative mandrel material. PVA was selected because it could be 3D printed and removed using water-based dissolution. Initial experimentation focused on evaluating PVA solubility, mandrel geometry, surface texture, and dissolution methods. A finalized injection molding workflow was developed using capped PVA mandrels, PVA-based mold release, desiccant storage, and ultrasonic bath dissolution. Silicone models fabricated using this workflow showed relatively consistent geometry and low residual material, but luminal texture and scaffold softness remained as limitations. The second aim of this thesis explored wax as an alternative mandrel material that was removable through heat application. Paraffin wax and water-soluble wax were initially considered, but beeswax was selected for continued development because it was less brittle and easier to handle during mold removal. A finalized beeswax workflow was developed that used a supported mandrel assembly, two injection mold systems, heat-based wax removal, and post-processing rinses. Silicone models that were produced with wax mandrels showed improved geometric consistency and shorter post-processing times, however residual wax remained a major limitation that reduced scaffold transparency and affected biological performance. The third aim of this thesis compared the finalized PVA and beeswax workflows based on fabrication practicality, concentricity, luminal texture, residual material, compatibility with the BVM setup protocol, and biological performance. The beeswax workflow produced models with stronger geometric consistency and faster mandrel removal, while the PVA workflow produced cleaner luminal surfaces and supported more favorable endothelial cell attachment. Neither in-house workflow fully matched the outsourced silicone model that each was compared to in terms of transparency, stiffness, fitting stability, and overall BVM compatibility. Overall, this thesis demonstrated that both PVA and beeswax can be used as alternative sacrificial mandrel materials for in-house silicone vascular model fabrication. Further, future work should prioritize refinement of the PVA workflow to improve scaffold stiffness, fitting stability, luminal smoothness, and reproducibility so that in-house silicone vascular models can more closely match the performance of outsourced silicone scaffolds.