Building a Large Format 3D Printer – Part 1: Pellet Extruder

Video Statistics and Information

Video
Captions Word Cloud
Reddit Comments
Captions
hey what's up guys it's dr d flow and this 4x4 foot aluminum plate is going to be the bed for my next 3d printer build once finished this printer is going to be capable of creating furniture and other ridiculously sized objects so stay tuned dr d flow a normal extruder like this one is not going to suffice what is going to be a very large 3d printer fortunately i was able to partner with massive dimension to bring in their md ph2 pellet extruder this extruder can print the equivalent of a roll of filament every hour force feeding filament in this e3d v6 hot end would require about 15 hours to run through an entire spool i'm going to dive into what allows this extruder to have 15 times the volumetric flow rate in a second but first we need to talk about feedstock the md ph2 extruder is fed with pellets not filament it's worth quickly talking about the relationship between pellets and filament plastic as a raw material is polymerized in a reactor where the resulting product is chopped up into small granules or pellets to create 1.75 millimeter filament for your ender 3 or similar 3d printer the pellets need to be heated up and extruded continuously through a 1.75 millimeter die i briefly demonstrated this process in my how to build a 3d printer video but the takeaway is that this extra step means that a roll of filament is going to cost more than the equivalent weight in raw pellets the pelleted extruder sidesteps the filament and takes the pellets directly which does equate to significant cost savings if you can purchase pellets in large enough amounts here is 20 kilograms of pla pellets which i purchased for about nine dollars a kilogram that is roughly a 50 saving over a spool of filament and additional savings can be had if you purchase pellets by the pallet which is hundreds of kilograms let's take a quick tour of this pellet extruder pellets will be fed down this clear tube where they will encounter a screw the screw rotates and transports the pellets down into the heated barrel not only does the screw transport the pellets but it also mixes them as they melt this mixing action allows for the polymer to be heated and melted more evenly the importance of this mixing cannot be understated when printing large volumes quickly you can't just crank up the temperature of the barrel and hope that the heat will conduct through all the pellets because you will burn the plastic near the inside walls as more and more pellets are moved into the barrel it gets harder and harder to turn the screw therefore a massive servo motor is used eventually the pressure will be high enough in the barrel to extrude molten plastic out of the nozzle if we zoom in on the nozzle the diameter is 1.5 millimeters massive dimension sends four nozzles all the same size when you purchase the extruder the extruder can support up to a five millimeter nozzle which you can create by drilling out the hole on one of the standard nozzles five millimeters may not sound large but for 3d printing it's comparatively huge for reference here's the standard 0.4 millimeter nozzle that most filament extruders use with that tour and background on pellet extrusion out of the way let's go ahead and melt some pellets i've hooked the pallet extruder up to the duet 36hc and this is going to serve as the controller board for this printer the duet 3 is the big brother to the duet 2 wi-fi which i recently used in my voron 2.4 build but the duet 3's expandability is pretty much endless this 1xd expansion board provides step direction and enable outputs for the servomotor on the extruder the duet 3 does not come with a wi-fi module so i'm using a raspberry pi to control the duet and host the web application for configuring and running this printer i will talk about all these details once we get to the wiring portion of this build but the last big thing is that the pellet extruder uses ac heaters so the duet is using this solid state relay to switch them on and off to regulate temperature the ac power allows the extruder to quickly heat up to a temperature of 450 degrees celsius it's worth repeating that max temperature 450 degrees celsius this means that the extruder can process polycarbonate with ease but can also print engineered grade plastic such as pei and peak in addition to a high powered extruder these polymers also require a heated chamber to print which would complicate the design of this printer but let me know in the comments if you want to see performance plastic printing the pellet extruder has come to temperature that extra noise are the cooling fans preventing the pellets from melting too soon but we can now add pellets into the hopper i followed the manufacturer's instructions and i heated these pellets for four hours at six degrees celsius to remove any kind of moisture any water present in the pellets is going to cause bubbles and voids to form in the extruded fiber i just hit extrude and you can barely hear the sound of the servo motor turning over the fan noise this thing is incredibly quiet it's going to take a little bit of time for the pellets to move from the hopper down the nozzle so i want to talk a little bit more about this technology an advantage of pellet extrusion is that you don't have to worry about the amount of filament on a spool and possibly having to change that filament mid print you just have to keep adding pellets now the amount of pellets in this tube is definitely not enough for printing large objects so i need to figure out a way to get a constant stream of pellets coming to the extruder while it's moving i'm thinking of suspending a large bag above the printer with a flexible tube that kind of follows this around i'm not sure yet we're going to get to that later here are first bits of plastic being extruded very exciting with my current extrusion speed the servo motor is only turning at 250 rpm i suspect that we can get much higher flow rates we're going to investigate that when we print that's pretty slick check out how it's coiling down here at the bottom that transparent material is cool what would be even more awesome is dyeing that extruded fiber with a colorant these little bags are enough to dye an entire kilogram of filament each let's try out the orange color and i think it's really going to pop the color change is not going to be immediate there's a lot of material that's held up in the barrel that will need to be extruded before the colorant can get there this is actually a good time to mention the main drawback of held extrusion having a large amount of molten material near the nozzle means that even when the servo motor stops turning the screw the plastic is going to ooze out of the nozzle there are tricks to minimize this ooze like feeding less pellets into the extruder moments before the extrusion needs to be stopped but this only helps so much when using a pellet extruder on a 3d printer you need to use paths that minimize retractions we will get into this more when we start printing and there may be some solutions that we can engineer such as a flap that quickly covers nozzle when moving to disconnected sections i think you can appreciate how much plastic is just kind of held within this whether it's actually inside the barrel or in some other kind of tubes but before i added the colorant and regular pellets together i removed all the previous pellets this tube was basically full when we started and we still haven't dyed the extruded fiber really at all don't make a liar out of me please turn colors oh finally i can see it wow that's a bright orange i like that color i wonder how much darker it's going to get it's going back to a slightly lighter color we need more of that colorant to build up in the barrel there it goes darker again eventually once enough color it reaches this lower part of the barrel combined with that mixing action of the screw you should get a pretty even colored fiber we'll go ahead and stop it that color looks pretty even that's awesome my first pallet extrusion print to get everything to work so smoothly i did have to make a small change to the extruder this k-type thermocouple was the stock temperature sensor and it is capable of reading temperatures far higher than the 450c limit of the extruder it was bolted to the back of the barrel which is what you're looking at now with the insulation removed the problem with this thermocouple is that the temperature sensing phase is conductive with the leads that go to the microcontroller [Music] this is not a big deal if you use the heater relay that comes with the extruder but i want the temperature to be controlled by the duet main board because the probe face is electrically conductive any noise near the pellet head will be picked up by the sensor which will give erroneous readings at the duet to be safe the duo will throw a fault and shut down the extruder whenever unexpected temperature jumps are recorded here you can see that when i click extrude the noise from the motor was enough for the duet to throw a fault the solution was to switch out that stock thermocouple with a temperature probe that comes in a three millimeter cartridge like this one because for the most part these are electrically isolated from the surface that they are sensing i milled a small adapter to attach this three millimeter cartridge to where the old thermocouple was mounted i drilled two holes so that i can mount two separate thermistors a 104 gt that is good for up to 300 c and a high temperature td 500 that can measure up to 500 c i currently only have the 104 gt but the td 500 is on order my plans right now are do all my testing with pla and abs so i'm not in a rush to get a high temperature probe this is the smallest part i've machined on the mill to date the adapter was designed to match the contour of the barrel to get good heat transfer for accurate temperature readings [Music] with the extruder all set we can talk about building a motion platform capable of moving this beast around typically pellet extruders are attached to robotic arms that provide a lot of flexibility in not only the printing volume but also the printing angle however robotic arms that can move this extruder which weighs about 18 pounds or 8 kilograms in a big enough work envelope are prohibitively expensive also the kinematics or the way the robotic arm moves makes it difficult to program and often requires proprietary software my goal is to have the same workflow for creating a printable file on a normal filament printer for this palette printer for this reason i decided on a cartesian style system where actuators will move the pelt extruder linearly along the x y and z axes i elected to go with 1500 millimeter c beam actuators from open builds the large cross-sectional area of the c beam creates a rigid actuator that won't sag during unsupported runs the lead screw will easily be able to move the pelt extruder around with the help of this high torque nema 23 motor the max speed of the actuator is 8 000 millimeters per minute or about 300 inches per minute while this is slower than a timing belt setup the acceleration of the printhead is more of a bottleneck than speed even with this beefy stepper motor it is unrealistic to reach these max speeds especially during the printing process where the printhead will constantly be switching directions so the ability of the lead screw to move large masses easily is far better than the speed of a timing belt [Music] [Applause] [Music] now you would be hard-pressed to find an actuator capable of this large of a travel for only a couple hundred dollars if you've seen my past videos and you know that i have built plasma tables routers and filament-based 3d printers all from open builds components before i ever started collaborating with open builds this price per performance was why i always elected to use their components but on a printer this large even cost effective linear actuators start to add up so big thanks to open builds for supplying the mechanical components for this build let's go ahead and look at my cad model to see how i've arranged everything this is going to be a stationary bed design where the extruder will move in all directions there are four vertical actuators found in the corners and these will move the gantry up and down the y-axis movement or the forward and backwards motion is handled by two linear actuators that connect to the z-axis actuators at their ends the x-axis that directly moves the pellet extruder spans the distance between the pair of y actuators it's very important that the z-axis columns are perfectly vertical if they're leaning in or away from the printer the gantry will bind as it moves up and down not good these ribs at the top and bottom will tie the actuators together and hold them at the correct angle for the ribs to do their job correctly they need to form a perfect square there are many ways to connect aluminum extrusion at right angles but i have elected to go with a custom corner piece that i machined on the mill the machine surfaces of these corners make it easy to line everything up i've added some extra tapped holes on the sides in the top and bottom which will be used for enclosing the build volume and possibly adding additional support to the frame if needed you may have noticed that these four perimeter faces look dull compared to the top and bottom ones this is because of the machining strategy that i used i wanted to cut five of the faces in one setup which is the easiest way to produce a square block because it only takes one more operation to remove the excess material from the back side however the depth of these corner pieces required a long cutting tool and my bench top mill is just not rigid enough to prevent such a tool from chattering or vibrating this is purely aesthetic as i can't feel any ridges or valleys with my finger once i flip the part over and remove the excess material from the back side i was able to confirm that the corner was square with eight corners complete we can put together the ribs [Music] you would think that building a large format 3d printer would be a great first project for my new shop but i'm still getting everything organized over there also with how long 3d printing takes it'll be much more convenient to pop down the garage to check on prints during initial calibration [Music] [Music] these 40 by 40 extrusions maintain the spacing between the ribs now you can get a sense of how large this printer is going to be after making sure everything was square this frame is ready for the actuators but i will not be installing those until the next video i'm waiting on several components to arrive that i ordered after making a small design tweak so definitely subscribe to catch the rest of this build if you have not already but before i let you go i want to make some progress on the build plate because i need to finalize my design before placing an order for a custom bed heater on a printer this large having a level bed is critical any unevenness in the build plate will be amplified across the four foot length resulting in poor bed adhesion or the nozzle gouging the build plate it doesn't help that my quarter inch aluminum plate is slightly warped and still has mill scale on it if i had unlimited funds i would purchase a massive precision ground plate but in today's economy even this imperfect plate cost a couple hundred dollars so we gotta do what we gotta do the plan is to use these springs to create a 25 point bed leveling system to make every inch of this plate parallel with the gantry in order for this to work a bolt will run through a countersunk hole in the aluminum plate and then go through this spring before tightening into a threaded hole present on a second plate tightening the bolt will lower the bed while loosening it will increase its height this is how the bed leveling works on most hobby 3d printers just at a much larger scale in addition to cutting 25 equally spaced holes into the aluminum build plate i also need to create that second plate for the bolts to tighten into let's take another look at the cab model underneath the aluminum plate there is one large square in a skirt that covers the perimeter the skirt is more for aesthetics while the middle square piece has holes to mount to the aluminum extrusion underneath it as well as these groups of five holes in the shape of an x a separate part will be attached these holes which will hold the spring in place more on this in a second to create the main plate as well as the skirts i turned to my new sidekick 8 plasma table if you've not seen the video on installing the cnc table in my shop then definitely check that out because the size and capabilities of this table are going to enable far larger projects than this massive 3d printer these plates contain a lot of small holes and when it comes to plasma cutting the rule of thumb is to not cut holes with a diameter smaller than two times the thickness of the stock material the holes that i need to cut are 5.5 millimeters and 4.2 millimeters which do not meet this rule because this 11 gauge sheet is almost three millimeters thick so i created a program that would mark these holes by firing the torch at their location for a quarter of a second this will pierce the material and act as a center punch for when i go and drill them out later the accuracy of this table will put those holes precisely where they need to be you do have to be careful with this approach to not fry the torch for too long because the mild steel can be heat hardened making it impossible to drill out those holes later i'm kicking myself a little because shop saver offers a drill add-on for this very application and when i was speccing out my table i was so close to pulling the trigger on it but i didn't before cutting the material i lowered the water level in the table just as a test there's always a trade-off between smoke control and finished cut quality a higher water level captures more smoke but does cause slag to get caught on the backs of the material this is because the molten material that is flung during the cutting process is quenched when it hits the water if the water level is near the back of the stock material then the molten metal won't make it very far before cooling down this water level test is why cutting the perimeter of the skirts was so smoky there was virtually no slag on the part but a bit too much smoke for the shop [Music] i loaded up the aluminum plate and pierced 25 holes in that the original plan was to tap all the holes in the x for m5 thread the middle hole would be for the bolt that runs through the spring and the four holes on the perimeter would be for a 3d printed piece that will hold the spring in place it would look something like this however i would only be able to get a couple threads into this 11 gauge material and if i strip the threads then i would be out of luck instead i had the mill churn out this aluminum version of the spring holder that has the threading incorporated in the piece so instead of drilling and tapping 125 holes in the plate i had the mill do it automatically i need 25 of these spring holders and i was limited by my stock material on hand so i had to pack the parts really close together and find an alternative way to separate them from each other come on so for the plate i just had to drill 5.5 millimeter through holes so that i can attach the aluminum spring holders [Music] to get this plate nice and straight andy and i are going to attach these four seed beeps [Music] [Music] so [Music] so this turned out way better than expected those spring holders are awesome by myself i was able to lay this plate down and get almost all the springs to line up let's talk a little bit more about those springs so the springs are an inch long and they're rated for 25 pounds per inch this is plenty stiff enough for this application you can see it kind of compresses when i sit down on it but that's because i haven't tightened down any of these bolts at all when i do tighten them down this thing will get stiffer but we really don't need that much force to manipulate this aluminum plate here i'm touching the corner with my foot and you can see i'm flexing that corner and the rest of the plate remains relatively stationary i'm probably going to complain later when i have to adjust all 25 screws to get the perfect level build plate but i'm happy right now it's simple and it seems to work let's go ahead and put the frame around it [Music] so this is still very much just a 3d printer skeleton but we accomplished a lot in this video we showed that the pellet extruder works with the duet 3. we have a nice square frame and probably the biggest accomplishment is this massive 4x4 foot build plate that features a 25 point bed leveling system in the next video we're going to install a bed heater as well as all of the linear actuators so we could be printing very soon 2022 is shaping up to be a crazy year for the dr d flow channel it would be great to have you guys along for the ride and with that i'll catch you in the next one [Music] you
Info
Channel: Dr. D-Flo
Views: 18,162
Rating: undefined out of 5
Keywords:
Id: m7JpumMS0Po
Channel Id: undefined
Length: 26min 10sec (1570 seconds)
Published: Tue Dec 14 2021
Related Videos
Note
Please note that this website is currently a work in progress! Lots of interesting data and statistics to come.