Senior Capstone
Improved Design and Manufacture of Filtering Media for Water Treatment Systems
This project was completed as the senior capstone of my undergraduate degree. The team working on this project was composed of myself, Vincent Cheung, Tyler Smith, and Julianna Hill.
Our client, Aquaculture Systems Technologies, LLC. (AST) of Baton Rouge, Louisiana, currently uses an enhanced plastic bead to filter and treat water. These beads are small and can sometimes escape with the filtered waste during washing, introducing plastic waste into the environment. They have a relative density of only 0.91 - 0.93 g/cc, which limits the packing and filtration ability of the beads. Our goal for this project was to develop an improved type of bead that is larger and less dense than the current design, and also contains geometric features that allow for biofilm growth necessary for filtering. Additionally, we had to design a scalable manufacturing method for the beads and provide proof of concept either by implementing the method at BU or through comprehensive modeling.
Our specific aims were as follows:
Increase the size of the beads to larger than ¼ in all dimensions to prevent escape from filter mesh
Lower the density of the beads to 0.8-0.85 g/mL
Include surface features to protect a 200-500 micron layer of biofilm
Attain a bulk porosity of 50-55%
Develop a scalable manufacturing process for the redesigned beads
Maintain a production cost (stock and processing) less than $1.50/lb
Final beads produced in this project
Bead Design Iterations
Main design considerations:
Protected Surface Area - surface area of regions not exposed to contact with other beads when they are tossed around during cleaning
Required for bacteria biofilm to develop and be protected during backwash
The target for the protected surface area was approximated to be 50% of the total bead area
Porosity - bulk property that determines the beads’ ability to filter waste, amount of space in a set volume that is not covered by beads
Target porosity is 50%
High porosity correlates with improved filtration
Design for Manufacturability
Focused on exploring external features that would be achievable using processes with the capability for scaling
We finalised the project with two designs: tooth and cog beads. The original designs for these beads were chosen as they followed the requirements set by the customer in simulations, and then were iterated upon to better their properties.
Tooth Bead
During testing of the original design for biofilm growth, it was noted that the gaps were too small. These became full of biofilm and had a decreased porosity, thus making them less effective at filtering since water could not pass through the now full gaps.
Iteration led to larger grooves that were more than twice the size of the desired biofilm (500 microns) to prevent a notable reduction in porosity.
Design for manufacturability: these were designed with the thought of making them with a continuous stamping and extrusion process, thus having geometry which could be made by pressing extruded material with two rollers.
Cog Bead
These gaps were too small to accommodate both biofilm growth. Thus, the gaps were widened and the number of these decreased to maintain the overall size. In addition, the gaps were designed in a way that prevents multiple beads from slotting together and causing a decrease in porosity.
Design for manufacturability: these were designed for an extrusion process, so the geometry could be made with a single nozzle.
Manufacturing Process
A scalable manufacturing method was developed and tested to make beads out of LDPE. Both beads used an extrusion method, with the tooth bead setup having the addition of rollers to stamp the design geometry. The process was fully manual, with the LPDE rod being pushed by hand through the extrusion setup and the rollers being cranked manually.
Cog Bead
The setup shown in the image was used to make a long rod of beads, which was then cut into individual pieces.
The nozzle was made using a wire EDM to create the final geometry desired along with the sloping walls to make the process of extrusion easier. The outer setup was made using 80/20 frames and steel parts manufactured in a CNC machine to maintain the nozzle in place.
Tooth Bead
The setup shown in the image was used to make a long strand of beads, and the beads were then individually cut out.
The nozzle used for this setup was featureless, with its purpose being only to heat up the LDPE for stamping. The rollers for stamping were made using SLA printing using Formlabs’ High Temp Resin to resist the temperatures of the LDPE pushing through.
Manual use: the manual pushing of the LDPE through the nozzle and manual rotation of the rollers as opposed to an automated system caused there to be variations in the size and definition of the beads.
Rollers: the rollers were not sharp enough to cut the beads as they were stamped, thus needing post-processing and causing variations in the size of the beads due to human error.
Limitations
Testing
To simulate bead properties prior to manufacture, the team used SOLIDWORKS to determine bulk properties such as porosity as well as single-bead properties like density. This allowed us to iterate on our designs and determine whether they would theoretically fulfill the client requirements or not.
The following process was followed to test bead properties:
The individual bead model was assigned LDPE as its material (setting the density to 0.94 g/mL), and its volume, surface area, protected surface area, and mass were determined. Size dimensions were also determined using the individual bead model.
The bead model was arranged into an array in which the beads are orderly packed to take up as much space as possible within a container of known volume. For this step, the beads were either packed all in a vertical or horizontal orientation, depending on which one would allow for the most volume to be taken up by the models.
The number of beads within the container was found, and the individual bead volume found previously was multiplied by said number. This value was then divided by the total container volume to determine the porosity.
This provided us with the “worst-case” scenario, as the beads taking up as much space as possible within a container would reduce the porosity overall due to a decrease in the gaps available for water to pass through.
SOLIDWORKS Simulations
Implementing a ‘Poor Man’s Filter’ to Test Biofilm Growth
To simulate AST’s filters and test whether our bead designs could grow and maintain a biofilm, we were provided with a ‘poor man’s filter’ design to implement. Below is an image of the design and our version, along with a video to demonstrate how it functions. We added fertiliser with ammonium chloride and sodium nitrite in the casing around the jar to act as the contaminants. In addition, we also added sugar to speed up the process of the biofilm growth. After this, we let it sit for multiple weeks in order to assess the growth on the beads.
The figure below shows the final bead designs before and after testing in the filter. The beads on the right were 3D-printed versions (printed in foaming PLA to simulate LDPE properties) which were used to test bead properties prior to manufacturing with LDPE. The beads on the left are the final LDPE versions. These were tested for a limited time due to fulfill project deliverable deadlines.
