The build process for CB2A Bioreactor involved implementing designs that met our requirements outlined in the previous page. We primarily used polylactic acid (PLA) to manufacture custom parts and used an Arduino to test the first few iterations. We also performed simulations in SolidWorks to explore shear stresses in silico.
Build Process
Computer-Aided Designs
MK. 1
A multipurpose grid was designed to house all electronic components, preventing contact with water. A 3D-printed mount (left corner) secures the peristaltic pump, the holder (centre) supports the NEMA-17 stepper motor, and two trays hold the microcontrollers. Bottom extrusions allow the parts to slide onto the grid and lock into place.
Figure 2. Physical prototype and 2D visual demonstration of caulobacter bioreactor
Electrical Circuit
Figure 3. Circuit diagram of the caulobacter bioreactor containing stepper motor for agitation, 12 V power adapter, DRV 8825 motor driver, 12 V servo motor, and temperature probe (breadboard)
Computational Fluid Dynamic Modelling
Rationale
2D culture can give important insights into the effectiveness of electroporation and the growth efficiency of engineered cells. However, in many contexts, the ultimate goal is to increase product yield so that it can later be purified and used in downstream applications.
In our CB2A bioreactor, agitation is a central design challenge. Proper mixing ensures that nutrients and oxygen are evenly distributed while preventing sedimentation. At the same time, excessive shear forces can damage cells. Our goal was to identify an agitation method that maximizes mixing efficiency while minimizing shear stresson Caulobacter crescentus. To address this, we used Computational Fluid Dynamics (CFD). Before committing physical and financial resources to building a prototype, simulations provided a cost-effective way to compare different mixing strategies. CFD allowed us to visualize fluid flow, estimate mixing times, and predict shear stresses under different conditions. Specifically, we tested three of the most common mixing approaches, axial, tangential, and radial impellers,by designing three impeller geometries using CAD and simulating their performance in silico. These simulations gave us a first look at which designs and parameters might provide the optimal balance between rapid homogenization and cell protection.
Ultimately, our research question was simple but fundamental:
Which impeller design provides the most effective mixing for our Caulobacter bioreactor?
Impeller Types and Mixing Strategies
Impellers are the rotating blades inside a bioreactor that drive fluid motion and mixing. Different designs produce distinct flow patterns, which in turn affect how quickly nutrients and gases are distributed, as well as the level of shear stress experienced by cells.
Axial Flow Impellers (a)
Radial Flow Impellers (b)
Tangential Flow Impellers (c)
Push fluid parallel to the impeller shaft (upward or downward).
Push fluid perpendicular to the shaft, toward the walls of the vessel.
Push fluid in a circular pattern around the vessel walls.
Create strong circulation loops that move fluid from the top to the bottom of the vessel.
Generate high shear zones and strong turbulence near the impeller blades.
Encourage rotational flow with lower shear compared to radial impellers.
Often used when rapid bulk mixing is required.
Useful for gas dispersion but can be damaging to shear-sensitive cells.
Can be effective when gentle but consistent mixing is needed.
By comparing these three designs, we aimed to understand how flow direction influences overall mixing time and the stresses applied to our engineered Caulobacter cells. Figure taken from [1].
Definition of a rotating region around the impeller to simulate agitation effects.
Figure 5. Definition of a rotating region (in pale blue)
Flow Simulation Wizard Setup
Launch the Flow Simulation Wizard to initialize the analysis.
Select time-dependent internal analysis, defining total simulation duration and time step resolution. ([2])
Select Sliding Rotation Type, This includes activating the time-dependent analysis.
Define a duplicate of the water fluid to represent the dyed phase, and assign its initial concentration in the region to 1.
Figure 6. Definition of the dyed water phase (in blue)
Fluid Properties & Separation
Specify both fluids for mixing (water and dyed water-like liquid) and ensure they exist in the same liquid state. ([3])
Define initial conditions with each fluid occupying distinct regions of the tank. Assign the initial condition water concentration to be 1. This will ensure that the two phases, dyed and non dyed, are separate.
Boundary & Rotating Region Conditions
Apply boundary conditions appropriate to your model, such as adiabatic walls or no-slip surfaces.
Goals & Outputs
Set simulation goals:
Average velocity
Volume fraction of fluid A/B over time
Torque or power on the rotating agitator These enable tracking of mixing progress and energy requirements.
Running Simulation & Postprocessing
Run the transient simulation and monitor goal convergence.
Explore results via the Transient Explorer to play through the simulation timeline or inspect specific frames.
Visualize:
Volume fraction plots showing how fluids blend over time.
Velocity fields, circulation patterns, and any regions of high shear or poor mixing
SolidWorks Simulations
Results - Flow Patterns
Axial Flow Pattern
Radial Flow Pattern
Tangential Flow Pattern
Magnetic Stirrer Flow Pattern
Results - Qualitative Axial Mixing
Axial - clockwise - 150 RPM
Axial - clockwise - 300 RPM
Axial - clockwise - 600 RPM
Axial - counterclockwise - 150 RPM
Axial - counterclockwise - 300 RPM
Axial - counterclockwise - 600 RPM
Results - Qualitative Radial Mixing
Radial - 150 RPM
Radial - 300 RPM
Radial - 600 RPM
Results - Qualitative Tangential Mixing
Tangential - 150 RPM
Tangential - 300 RPM
Tangential - 600 RPM
Results - Qualitative Control Mixing
Control- 150 RPM
Control - 300 RPM
Control - 600 RPM
Discussion
Our simulations provided several key insights into the mixing dynamics of the CB2A bioreactor. First, we observed a clear relationship between impeller speed and mixing performance: faster rotation was consistently correlated with shorter mixing times. This result is intuitive, as increased angular velocity enhances fluid circulation and turbulence, leading to faster homogenization. However, this benefit must be balanced against the potential for increased shear stress, which can negatively impact Caulobacter crescentus growth and viability.
A second important observation was the role of impeller surface area in determining mixing efficiency. Designs with a larger blade footprint were able to move greater volumes of fluid per rotation, promoting more effective circulation throughout the vessel. This suggests that impeller geometry, beyond just rotational speed, plays a decisive role in achieving efficient mixing.
When comparing designs, the anchor impeller showed the most promise for our application. Its geometry promoted strong bulk circulation and minimized dead zones, while avoiding the extreme shear stresses associated with more aggressive radial impellers. These properties make it particularly well suited for culturing Caulobacter, which requires homogeneous conditions but may be sensitive to high shear environments.
While CFD provides powerful predictive insights, our findings remain preliminary. To validate these results, experimental testing will be essential. Future work will include growth curve experiments, survival analyses, and direct observation of mixing behavior in physical prototypes. This will allow us to confirm whether the predicted benefits of the anchor impeller translate into improved cell health and productivity in real cultures.
Limitations
Although our CFD simulations provided valuable insights into the mixing performance of different impeller designs, several limitations must be acknowledged:
Simplified Fluid Properties We modeled the culture medium as water, assuming Newtonian behavior. In reality, microbial cultures often have non-Newtonian characteristics (e.g., viscosity changes with cell density), which can significantly influence mixing and shear stress.
No Gas Exchange Modeling Our simulations did not account for aeration or oxygen transfer, both of which are critical for microbial growth in bioreactors. Gas—liquid interactions can change flow dynamics and impact mixing efficiency.
Boundary Condition Simplifications The vessel walls were treated as ideal, and no baffles were included. In real bioreactors, wall roughness, baffles, and fittings strongly affect flow patterns.
Focus on Mixing Metrics Only While we quantified mixing efficiency and shear distribution, other factors relevant to cell culture, such as oxygen distribution, nutrient gradients, and energy consumption, were not explicitly studied.
Qualitative Nature of Results The simulations provide qualitative predictions of flow patterns and mixing performance. Quantitative accuracy may be limited due to model simplifications, and the exact mixing times or shear rates may differ in a physical system.
Validation Pending CFD is a predictive tool and cannot fully replace experimental validation. The actual performance of the impellers must be confirmed through growth curves, survival analysis, and prototype testing with Caulobacter crescentus.
