 Hi everyone, today me, Victoria Haricunyan and Nae Vartanyan are going to present our capstone project, which is a 3D printing filament recycler. As you can see, here is the device. Our supervisor is Arman Asadyan and our advisor is Sarkis Ektunyan. So our content includes introduction, project overview, methodology, results, conclusion, and then references. So the main problem that our project aims to address is the considerable amount of 3D printing, considerable amount of waste, filament waste that 3D printing produces, which makes environmental issues, resource depletion, and higher expenses. And also we ourselves encountered this problem in our university labs and other maker places. So that motivated us to create this project. Okay, so these are the three main modules of our project. And this is the extruder, which is the main and the most complex part. And it is responsible for the melting and extrusion process of the filament. So we add the pellets from here, the hopper, and the extrusion takes place from here. And then we have the cooler module right here. And whenever the filament is extruded, it is still in a biscoelastic state, so it's like a gum. So we pass the filament through these holes manually, and it is cooled down and solidified during this process. And then we have a spooler, which we made it manually. And yeah, we didn't bring this part, because we couldn't pass it anywhere. But we have the spool itself. And it is for wrapping the filament after it is extruded and passed through the cooling part. So let's talk about our methodology. Our methodology includes two parts, research and implementation. We conducted research about different plastic recycling, different types of plastic recycling. Also working principles of industrial filament recyclers. And for component selection, we did research based on compatibility, safety features, and cost. And our implementation included creation of 3D model and manufacturing, technical specs, connections, and experimentation. So let's talk about 3D models and manufacturing. First of all, let's talk about this frame. The models were created with SOLIDWORKS software. And now you can see the parts from the picture and from here as well. So these upper and lower corners provide support, and they were 3D printed and modeled using SOLIDWORKS software again. Also we have here these mounting plates and these ribs that support the mounting plate. And in the mounting plate, the extruder and the motor are connected. We have here used, we used MDF base. It was CNC machine and these mounting plate and ribs were CNC machine as well. We use, and also yes, we use plexiglasses that were laser cut it. And we have here hopper. It is removable. It is 3D printed and these hopper base, you can see from the picture as well. It was CNC machine as well. So now let's talk about the main parts of the cooler module. We have mainly the cross flow fan here. We have the safety cover above it and the main frame that was designed in a way to provide good air circulation. And here there are holes that provide place for the filament to pass through them. So now let's talk about technical specifications of the extruder. As you can see our main part is this. It is not visible here because we insulated it with fiberglass material with the help of our friend Alex here. And insulation was needed because the barrel, you can see here, and these are the heaters. They produce high heat and we wanted to insulate the material in order to avoid overheating the electronic components here. So we have three heaters as I've already mentioned. The barrel, the screw side, the barrel. We have nozzle with 1.75 millimeter diameter. We used this timing belt pulleys with 3 to 1 ratio. We used NEMA 23 stepper motor. Here you can see the temperature controllers. We have here also the on-off switch that turns on and off our device. We have the potentiometer that controls the speed of our motor. We have these connectors to provide, to make our device portable and compact. And also we have here our AC male connector. Okay, so these are the main electrical components of our device, which are placed under the MDF base you may see here. So first we have three solid state relays here and these are controlled by the PID controllers. And they have two modes of normally open and normally closed, which are used to turn on and off the heaters. Then we used an Arduino microcontroller in order to write a program and driver stepper motor and the driver. And then we used this digital stepper driver, which allows us to control the high current NEMA stepper motor with using just a very low current microcontroller. And here we have our power supply, as you may see here too. It is a 24 volt DC power supply and it supplies most of the power needed for our electrical components. And then we have a 24 to 12 volt DC-DC converter, which converts the 24 volt to 12 volt and supplies it to our Arduino and also the cross flow fan. Okay, so here we have a diagram, power diagram, which we did based on the power that each component is being supplied. So first we have the 220 volt AC power, which is supplied to the rocker switch here. And from there it is supplied to the 24 volt DC power supply, the three PID controllers, the band heaters, which are under the fiberglass, and the three solid state relays. And then the 24 volt is converted to 12 volts in the DC-DC converter and supplies it to our Arduino and cross flow fan. And also the NEMA 23 stepper motor is supplied 24 volts through the stepper driver. So we focused mainly on the quality control. That's why we conducted filament quality control test. And the objective was to obtain a filament with high quality and filament can be qualified by the presence of bubbles, texture and ductility. Here you can see the results of our first experiment. We're going to talk about this now. I'm going to talk about the experimentation process. So the setup was, as it is shown here, with these temperature values because based on research we found that it is recommended to set the temperature values in a sequentially increasing order. So we did the same. However, as you can see, these are like the step temperatures and these are the actual temperatures that thermocoupled to test. And as you can see, especially in the middle temperature was way too high from the step temperature. And we thought that this might be because our barrel is a metallic and it is a good heat conductor. So two other heaters transferred the heat to the middle heaters, heating element. And that is why it is going way too high from the set temperature. However, these values caused, as you can see from the sample, caused bubbles in the filament and it has low ductility. It can easily be broken and it is of bad quality. Question, the PIDs just control the temperature? Yes. Okay. So in order to address the issue of the bad quality and also because the issue of the middle controller that was very high values, we decided to do another approach. So we set the value of the middle heat controller to a value lower than the room temperature, so that the relay connected to it would be at normally open mode at all times and the heater wouldn't heat up. And the, yeah, and as you can see from the samples that you have, the product is more ductile. That would break easily and there are no visible bubbles inside it. So, yeah, we got very good results. But we kind of, it was our own approach. It wasn't what was recommended. And we got good quality and different motor speeds and especially at 18 rpm. However, one issue that we faced, which was because of that piece heater, which is very close to the hopper, it has a high temperature. So when the amount of pellets that we would add to the barrel were not controlled, sometimes they would, it would be half melted and it would stick to the hopper base here and here and it would have difficulty moving forward. So in order to, yeah, to overcome this issue, we decided to, yeah, this is for the third experiment. So we decided to do a third experiment to lower the value of the heater, which is very close to our hopper. And we set it on 170 and it actually solved the issue of the sticky filament. However, because it had the temperatures not high enough, we had to lower the speed to 10 rpm so it would have enough time to melt. And however, as you can see from the samples you have, the quality is not as good as the previous one and there are some bubbles in it and it has a good ductility. However, it is, the surface is slightly more textured and it has a lower quality. Yeah, and these two are the best results we got and the worst. These are the best results and these are the worst results? It is more obvious on a small scale. So we can say based on the amount of experiments we were able to conduct that we reached the best quality when we had the heater temperatures as they follow and with the screw speed of 18 rpm. So based on our results, we did performance evaluation and with these temperatures and this screw speed of 18 rpm as we got that these are the most optimal ones and provide better results. So we had that the time needed for the heaters to set to these temperatures is approximately 18 minutes and the time it takes for the pellets, and yes I forgot to mention that we used these pellets from recycled ABS plastic and the time it takes for these pellets to be melt and extruded from the filament is approximately 2 minutes. And based on our calculations we got that 1 kilogram of filament can be obtained in 5.3 hours and 5.3 hours consumes 2.65 kilowatt power and also we know that 1 kilowatt of power is approximately 50 drums so we calculated the price for our filament and it is approximately for 1 kilogram we have 133 drums. So now let's get to the conclusion. Actually there are future works and improvements that can be done to make our device better and the most important one is adding an automatic puller module because our results ensured high quality but as we did this pulling and pulling manually we didn't get the constant results in the diameter so adding a puller module can solve this problem and the device can provide high quality filament with constant 1.75 millimeter diameter. Also a sensor can be incorporated with a PID controller inside the puller module to get feedbacks about the diameter of the filament and conducting more experiments with even different types of plastics we did only with ABS but it can be done as well with PLA as well and adding an LCD display or predefined molds based on the optimal results acquired from a sufficient amount of experiments can make our device more user friendly. Okay so this was a complex project and we faced many challenges so some of them were during the assembly process and some were during the experiment so one issue that we faced during the assembly was that we constantly had to make changes to the frame because we were either adding new elements on the surface or changing the layout so it kind of it was a lot of work and we kind of postponed our work from time to time and also we had issues we experienced issues with one of our heat controllers we thought we damaged it but fortunately it was just a loose wire connection which was detected by our friend Aram. I don't know where he is. So we also faced some challenges during the experiment. The first one was improper air ventilation in the lab which caused the smell of the filament to stay in the lab and it wasn't pleasant for the people who were working there so we couldn't conduct more experiments as we wanted and also we experienced some issues with the space because as you can see it kind of requires a lot of space to work properly and unfortunately because of lack of time we were unable to make an automatic puller module. However since this was a multidisciplinary project it demanded knowledge of mechatronics, cut, electrical and mechanical engineering so we had a lot of learning to do and so what we learned was we experimented a lot with designing and prototyping especially using the solid work tool and we learned how to work with equipment in the lab for example the 3D printers, the 3D scanner and the CNC machine and we also learned the working principles of the stepper motor and Arduino coating both during the capstone and the mechatronics course and also we gained a lot of practical skills such as assembling and doing the wiring. Okay so now no one wanted to sit here live because of the smell yet it was a popular demand that we do not demonstrate it here so we have a video. You're welcome. You should get used to the smell. We are used. This is the main process. That was all, thank you and also I have to, sorry, thank especially our supervisor for providing constant help and our friends for as you saw won't help with identifying an issue and other with the fiberglass and also carrying the device around for us and also I kept us with some issues we're experiencing with our code. Thank you. One question, have you done any printing? We were afraid to. We couldn't get the constant diameter, that's why we were afraid to experiment with our. Yeah because we didn't have the puller module which provides the constant diameter so we wouldn't get by manuals pulling we would get different diameters so we were thinking that extruder inside the printer wouldn't be able to catch the filament. Do you think, maybe this is a guess, do you think that if you had a slight vacuum inside it might pull the air bubbles out? Yeah. We saw a lot of 3D printed projects today. Have you collaborated with colleagues to recycle their materials? We weren't about that. The timing wasn't right. After the capstone we would collect everything else. My second question is that we all know that whenever you heat up and reshape the material it starts to lose its mechanical properties over time so as a rough estimate you have any idea on how many times you can recycle the same material under the same conditions until it becomes virtually impossible to use. We don't have a number, however we did research and we found out that ABS is a material that very hardly loses its initial mechanical properties and we did when we were extruding some filament. We were using it to recycle so it was like 3 or 4 times being recycled. But it is not recommended. I think 2 times is the maximum. Because we were using already recycled ABS pellets. But still, pre-tick is not recommended. We can risk and fix something. Last question. You mentioned about the future work which is related to this cooler. Except that what other aspects do you feel like? For example, Hermione having that cutting it all the time because most of the time it's not that shape. Are you thinking to adapt that to be a future work? Yes. These are already ready pellets. But for that we can incorporate a shredder and we have shredder and we can easily shred every weight. No, it wasn't shredded. No, we obtained the pellets in this form already. We also like... We didn't purchase it. No, we got it from the recycling company. But we also broke our filament and added it to them. But the shredder can be incorporated in order to have... It seems like if you would combine 3 capstone projects into this, it can become a business. That's the plan. After the graduation ceremony we will get together and start thinking about the business ideas. Our goal in fact is to help boost the scalable technology aspect to the epic labs anyway. And jokes aside, our goal for next year is to make these kind of capstone projects multi-disciplinary computer science students and data science students and business students together to do a bigger project. Thank you.