Team:Exeter/Hardware
Hardware
Overview
The aim of the project was to prevent the release of PET microfibres into the environment from the washing of polyester (PET) clothing. Plastic microfibres from clothing account for 35% of all microplastic pollution, a large proportion of which is PET. In order to carry this out this goal, a filtration unit was designed to be fitted to household washing machines , with an enzyme release feature that, when attached to the outlet of a washing machine traps the synthetic fibres and holds them in place until they are fully degraded by the enzymes. The filters, having been tested with a modern Miele washing machine, have demonstrated the ability to capture over 68% of all detectable microfibres released. These can then be contained within the filter until their complete enzymatic degradation.
Introduction
There is a need for a reduction in microplastic emission from washing machines, and to play a part in facilitating this reduction, the end goal for our project was to degrade polyester fibres using the enzymes PETase and MHETase, thus stopping them from ending up in landfill or aquatic environments, where they are being consumed by marine life. In order to achieve that goal it was key to design a practical method of isolating the microfibres from the water so that they could be brought in contact with the enzymes. Our initial research showed that each wash releases anywhere from 30L to 100L of water, therefore making it infeasible to store all the water released. This approach would also have raised further issues pertaining to the concentration and volume of enzyme solution required to have an adequate working concentration such large volumes of water. As such the engineering task was to design, build and test a water filter capable of capturing microfibres, and containing enzymes so that the fibres are broken down on site within the filter without having to store large volumes of contaminated water.
In order to structure our approach we followed Pugh’s Total Design theory, shown in the figure on the left, in the designing and manufacturing of the filter. This is a design method commonly used by engineers whilst developing new products, and it includes all the core design processes and principles necessary for a completed model. By following the progression in each step we ensured that nothing was overlooked and all work was carried out with a clear purpose.
Market Research
This product would be sold nationally to washing machine manufacturers so as to be installed at the manufacturing stage. There would be a large market, due to the advancement of technology within washing machines, and the development of smart washing machines, as well as the current trend of increased sustainability. The target market for this product would initially be high-end washing machine manufacturers who are more able to front the initial investment costs of integration. Once there is a more widespread use of the filter and lower production costs with time, the target market would increase to cover all washing machine manufacturers.
Currently, there are not many similar products available on the market, and none have the capability of breaking down the captured fibres. We have analysed several other competitors who produce filters to capture microplastics during a standard washing cycle. The distinct difference between our filter and these competitors we have investigated is that our filter will degrade the plastics that are caught, not just dispose of them to landfill. Microplastics in landfill inevitably get washed by rain back into the water system, so our filter avoids the risk of microplastics actually ending up in the ocean when no others can guarantee this. This will inevitably increase the cost of our filter in comparison with the alternative options, however degrading the plastics should reduce the number of replacement filters that are needed.
To see a more in-depth market analysis see our Entrepreneurship page.
Product Design Specification
The construction of a Product Design Specification (PDS) was done to focus the designs on solving the problem at hand in a way that meet the demands of the stakeholders we have contacted, and so that a viable product was made that could be brought to market.
| Design Aspect | Considerations and Constraints |
|---|---|
| 1. Performance |
|
| 2. Cost |
|
| 3. Environment |
|
| 4. Life in Service |
|
| 5. Maintenance |
|
| 6. Packaging |
|
| 7. Quantity |
|
| 8. Competitors |
|
| 9. Shipping |
|
| 10. Weight |
|
| 11. Aesthetics |
|
| 12. Geometry and Size |
|
| 13. Material |
|
| 14. Standards and Specifications |
|
| 15. Installation |
|
| 16. Testing |
|
| 17. Safety |
|
| 18. Customer |
|
| 19. Manufacturing Facility |
|
| 20. Quality and Reliability |
|
| 21. Shelf Life |
|
| 22. Patents |
|
| 23. Documentation |
|
| 24. Disposal |
|
The use of the PDS immediately ruled out some initial concepts, notably the one below as it infringed on existing IP and used a motorised system which washing machine company Miele dissuaded the use of, due to additional energy costs which would negatively affect the energy efficiency of the machine as a whole..
Conceptual Design
Concepts
Several concepts were created using the CAD software Solidworks, in order to meet the PDS outlined above. Click each design below to find out more about it.
A
Outline: The fiter support comes off with the top, using a simple screw mechanism. When closed, the filter sits above the outlet pipe and below both the enzyme and inlet pipes.
Pros: The benefit of this filter design is that the filter disk be more easily accessible to the user of the washing machine and therefore easier to clean manually when required and/or replace.
Cons: A flaw in this design when manufacturing at student level is that the filter disk must be sliced to allow the central spindle to go through. This causes issues as water can pass through the slit, this can be mitigated by glueing or taping it back together but even this has its faults.
B
Outline: This model benefits from a simple, user friendly design that is ideal for testing.
Pros: The major benefits from this are that there are only two parts involved, additionally the inline design, with the inlets on the top and outlet at the bottom make it ideal for prototyping as it operates with gravity. The simplicity gained from this means that this model is ideal for a testing prototype.
Cons: The downside to this design comes from the fact that it is harder to clear as only the top of the filter can be accessed and is permanently attached to the base. The model can also not sit flat on a surface and must be held up or supported.
C
Outline: This concept uses a basket design to maximise the surface area of filter. We cannot implement this one as it would require sheets of nylon mesh, however it could prove more suitable for use in the washing machine.
Pros: The benefit is that the basket can be completely removed and easily cleaned and/or replaced. The large surface area is also very beneficial as more microfibres can be trapped before clogging.
Cons: The downsides to this model come in the cost of the nylon mesh in order to cover the large surface area as well as the difficulty in 3D printing the model which makes it harder to use for rapid prototyping.
D
Outline: This is a modular design for which each part can be switched out, cleaned or remade with modifications. It follows a simple design with two inlets and can sit on a flat surface.
Pros: The modular nature means this model is very easy to use for prototyping as small changes can be made without having to make a whole new device. This model also had very few issues printing and worked well in practice.
Cons: The surface area is more limited than in other models and the use of three parts makes it harder to manufacture.
Design Selection Matrix
We analysed each of our concept designs and compared them to each other using a matrix analysis (shown below). This decision matrix was constructed from the most important points from our PDS. Each criterion was given a weighting based on its relative importance to the product and subsequently each concept was given a score between 1 and 10 in each category, with 10 being the highest score. The final score reflects the overall suitability of each concept.
This matrix analysis showed that the best filter concept was design B, as it had the highest overall score and a showed its benefits with regards to performance, cost and size.
| Weight | A | B | C | D | |
|---|---|---|---|---|---|
| Performance | 10 | 4 | 7 | 10 | 7 |
| Cost | 5 | 10 | 10 | 5 | 10 |
| Maintenance | 5 | 10 | 7 | 5 | 8 |
| Weight | 2 | 10 | 10 | 7 | 10 |
| Aesthetics | 1 | 10 | 5 | 5 | 10 |
| Size | 5 | 10 | 10 | 4 | 7 |
| Installation | 5 | 6 | 8 | 2 | 5 |
| Final Score | 250 | 280 | 199 | 250 |
Prototyping and Testing
In order to ensure a continuous feedback loop from our testing to our design we opted to 3D print our prototypes in order to carry out Rapid Prototyping. Each iteration on a model was tested and then modified following the test.
Our experimental setup consisted of washing two large 100% polyester fleeces in a standard 1 hour 24 minute setup, the flow rate and volume output were measured as the output water ran through the filters.
On the first few occasions there was excessive leakage and/or the filter broke under the high pressure. However, on subsequent iterations the microfibres were captured and the testing developed towards optimising flow rates and enabling the filter to be used for subsequent washes without the need for cleaning.
A major design iteration occurred when Miele provided us with their washing machine dosing system to use for testing. This is an automated system that controls the release of detergent and fabric softener in the washing machine so that the consumer doesn’t need to concern themselves with it. Upon discussions with Miele the thought was that the enzyme delivery system could be incorporated into the same system as the detergent and fabric softener to facilitate the filter’s integration into their machines. As such the models were changed to fit this system as it gets us closer to the Product Design Specification with regards to an integrated system within a washing machine. The design result is a smaller model with a smaller enzyme inlet pipe.
Testing Capture Effectiveness
To fulfil their primary purpose any filter we design must demonstrate its ability to capture microplastics, further to this it is also important to know what proportion of the microplastics are captured. To measure this, we used three different microsieves with pore sizes of 40, 70 & 100 micron in order to weigh the mass of microplastic captured in each filter mesh.
The same wash was set for each microsieve: two XL 100% polyester fleeces in a standard wash setting. The outlet from the machine fed into the filter prototype so that all water ejected by the machine must pass through it. The outlet from our filter then fed into XFiltra™, a high end microplastic filter prototype provided for use in our testing by Xeros Technology Group. Our prototype design is different from the design of X-Filtra™ due to our degradation system meaning that the filter does not need to capture as many fibres, along with our aims of minimising costs and optimising simplicity. The purpose of XFiltra™ was to capture microfibres not collected by our prototype. The prototype and XFiltra™ were dried in a 30-degree oven for 48 hours and before the contents of each were weighed.
"XFiltra™ has been developed by Xeros Technology Group as an efficient and cost-effective solution to the issue of microplastic fibre pollution resulting from laundry and the cleaning of textiles. It is designed to be scalabe and easily incorporated into any washing machine during manufacture. XFiltra™ is a combined filter, pump and de-watering unit capable of capturing up to 99 per cent of microplastic fibres generated during the laundry process."
The results from the 100-micron filter are shown above, our prototype was able to capture 68% of the microfibres whilst the remaining 32% shown were captured by XFiltra™, meaning they escaped our prototype. The 70-micron mesh captured 29mg and the 40-micron mesh captured 41mg, with the finer meshes the mass of the fibres not captured by the prototype was below 20mg (the threshold for the balance) and qualitatively appeared to represent about 10%.
Whilst not ideal, all three mesh sizes failed to capture all the plastic microfibres, we are confident in saying that we are capturing the majority of microfibres released, though any further work should look to improve this.
Testing Successive Washes
The filters need to be able to operate correctly for multiple successive washes in order to prove their viability as a product rather than a prospect. In order to test this, we washed four XL 100% polyester fleeces using synthetic setting on the washing machine. The outlet from the machine fed into the filter prototype so that all water ejected by the machine must pass through it. A 100 micron micro-sieve was chosen for this experiment, as shown above it was found to capture over two thirds of detectable microfibres.
| 1st Wash |
2nd Wash |
3rd Wash |
4th Wash |
5th Wash |
|
|---|---|---|---|---|---|
| Flow Rate 1 (L/min) | 11.9 | 12.4 | 12.1 | 12.2 | 12.1 |
| Flow Rate 2 (L/min) | 11.8 | 12.2 | 11.9 | 12.1 | 11.9 |
| Flow Rate 3 (L/min) | 11.5 | 11.2 | 11.7 | 11.8 | 11.5 |
| Mean Flow Rate (L/min) | 11.7 | 11.9 | 11.9 | 12.0 | 11.8 |
We carried out one wash per day, using the method described above, every day for 5 days. Additionally, we recorded the flow rates from each wash to see if they decreased as the filter filled up. At the end of the week we had collected a significant amount of microplastic, shown in the picture above, and had no issues of clogging or blockages.
Whilst it could not be tested, we anticipate the filter could withstand several more washes before any issues would occur. This conclusion comes from the fact that there is still more space in the filter as well as the fact that there was no decrease in flow rates across all 5 washes, as shown above.
Testing Flow Rates
Flow rate is an important measurement as it helps in understanding how much resistance the filter mesh is providing which correlates the strain it places on the washing machine pump.
We compared the microsieves with mesh sizes of 40, 70 and 100 microns by recording flow rates at the outlet of the filter throughout a standard wash containing two XL 100% polyester fleeces. This was done by placing a flowmeter to the outlet of our filter and recording the maximum flow rate from each water output.
As seen above there was no apparent difference in flow rates across any of the three mesh sizes.
Mechatronic Design
Mechatronics is a multidisciplinary branch of engineering which focuses both on mechanical and electronic engineering systems. We used these design principles to further realise the integrated unit aspect of our filter, an operational system that could be included in a washing machine. Our complete unit would have to be able to dispense our enzyme solution when enough microplastic fibres had been collected and maintain these submerged for the correct period of time to ensure full degradation had taken place.
To achieve this, our system would have to have a way of knowing how much microplastic build-up occurred on the filter mesh, currently this is estimated by way of a flowmeter but would need to be refined later on. The system would also need to know how much enzyme solution to apply and when to release this solution so as to be washed out with the next wash. It would need to have mechanical systems to administer the enzyme solution to the filter unit, stop water flowing into the filter unit and stop the enzyme solution flowing out of the filter unit before full plastic degradation has occurred.
To achieve these requirements we use multiple components:
- an Arduino Uno microcontroller
- a 4 channel relay module (5V)
- 2 normally ON Solenoid valves (12V DC)
- 1 normally OFF Solenoid valve (12V DC)
- multiple wires
- a breadboard
We used the normally OFF valve for the purpose of stopping the enzymes of flowing into the filter unit until required. The normally ON valves on both the inlet and outlet of the filter unit to allow contaminated water out of the washing machine to flow through the micro-mesh until a degradation was to take place. The figure shows the testing stage of the mechatronic design.
As well as these components we also had a WiFi module for the purpose of controlling the system through an online server, and a flowmeter which would allow the system to know how clogged the filter is, informing when to deliver the enzyme. These later components would be integrated if we had more time on the project.
Here are some useful links on how to build a similar control system using the same components for interested iGEM teams:
Relay Tutorial
Relay Tutorial 2
Electronic schematic
The schematic below is an example of how you may use a low voltage microcontroller such as an Arduino with a Relay Module to control higher voltage components [2]:
Components
| Arduino Uno | Relay Module | Solenoid Valve |
|---|---|---|
 |
 |
 |
Electronics Operation
Coded in the Arduino is a timed ON/OFF loop. This signal is sent to the Relay which in turn powers the solenoid valves, inducing a current in them, flipping them from their 'normal' state. This in practice and integrated with the filter and enzyme delivery system would be what stops water flowing into the filter unit and allows enzyme solution to flow onto the mesh holding the microplastic fibres. This code can be found on the Demonstrate page.
Software
When considering our entire integrated filtration system and the entrepreneurial goals we had set for our product, we decided that if our filtration system was to be a seamless part of washing machines sold to customers, there needed to be an element of user interaction. Google Firebase has supported our project and we have utilized their cloud infrastructure and services. We made use of various open source technologies, notably a Flutter template by Mitesh Chodvadiya. The prototype app provides vital information to the customer about the current state of our microplastic filter.
The app can be downloaded from the Google Play Store
Features of the app include:
- Counting the number of washes
- Measuring the approximate number of fibres degraded by using the filtration system
- The percentage of enzyme solution remaining
- A notification to signal when the user needs to refill the enzyme solution
- The status of the device
- Selection of data sharing preferences
The WiFi module integrated into the final filtration system would monitor how much enzyme solution has been passed into the filter, based on the number of washes. It will then communicate this information to the user via the interface, giving them the peace of mind about refilling the system.
With the app showing how many fibres have been degraded and the number of fish that have benefited from using the PETexe filter, it gives direct feedback to the user about the positive impacts they are having on the environment, showing them that they are making a real difference!
Architecture (Simplified)
The Architecture is the structure and relationship of hardware and software items to form the overall system. The Arduino microcontroller along with the ESP8266 WiFi module, feed data to Google Firebase. This module utilizes the Firebase ESP8266 Arduino Library. The app on the mobile device is integrated with Firebase and receives it on update by making use of Firebase Realtime Database.
Detailed Design
The final design was decided upon through the use of the design selection matrix, shown in the Conceptual Design section, in addition to many subsequent improvements that came with each iteration to improve the performance observed when testing each prototype.
An inline filter is common for any other filter types and has the principle benefit of simplicity above all else, the filter is able to behave as part of a pipe and as such can be positioned with relative ease anywhere in the washing machine system.
The most striking visual change from the first to last iteration of the filter is the final design enzyme delivery inlet, this was changed to final design to facilitate its adoption by enabling our filter to final design fit in line with washing machine manufacturers existing setups. The larger inlet worked for our initial valve design, however Miele then reached out to provide us with a dosing system which already exists in their machines, as such the valve was replaced with a smaller peristaltic pump which ‘doses’ the filter with an exact volume of enzyme solution.
Another significant, though less striking, change was the adoption of an O-ring to further seal the system. This change was provoked by slight leakage coming from the threaded intersection between the top and bottom halves of the filter at the extremely high water pressures generated. The O-ring successfully improved the situation and the prototypes with the O-rings included did not suffer from any leakage problems.
Our mechatronics design went from being a simple circuit, where we controlled the solenoid valves by coding it onto the Arduino board in a loop. Allowing us to ensure we could control enzyme release and water flow. To our final iteration which includes the ESP8266 Wifi module. This allows us to control the relay and hence the solenoid valves, from a Firebase server that has been designed for this purpose. This means that we can control the electronic components of the filter design remotely. We also have our app which is connected to our server, also allowing control of the solenoid valves, the enzyme delivery and the water flow from our app.
This design can and should be improved upon in order to further develop the product, however at present the final design is successful in removing microplastics from washing machine outlet water, can operate for several consecutive washes without the need for cleaning and shows characteristics required for its easy installation into washing machines thus fulfilling all functions demanded of it.
Manufacturing and Sales
Although currently still in the prototyping phase, we have looked into the future development and commercial production of our design for the use within household washing machines. PETexe would currently be aiming to sell to premium washing machine manufacturers, with the expectation that as the company develops and costs are reduced the filter would then be more accessible to a wider audience.
There is further potential for growth in other markets including fashion mills and water treatment centres in which there is a need for microplastic filters and relevant parties have expressed interest in pursuing this. A full operational plan presents the details relevant to production of the filters including product development, personnel, quality control, location and suppliers. Additionally, customer service, inventory control and the legal environment are outlined for the sale of the filters. The finances for the first year, including costs, cash flow, balance sheet and profit and loss statements are included and analysed. The forecasted finances show a net profit in the first year before tax with a net profit margin of 7.6%. After tax there is an expected loss, however this is forecasted to improve in subsequent years as production becomes continuous throughout the year.
Further information on our manufacturing and sales, along with a developed business plan, can be found on our Entrepreneurship page.
