Building a Selective Laser Sintering (SLS) 3D Printer!

[Music] [Music] dr d flow [Music] hey what's up guys it's dr d flo and man do i have an awesome build for you today in this box beside me is currently the most affordable sls 3d printer that you can buy well i guess technically it's the parts to make that printer because it comes as a kit i'm excited to get it unboxed but first if you're not familiar with the process of selective laser centering or sls for short then definitely check out the video i released alongside this one that explains everything you need to know about this unique additive manufacturing technology let's get into it i'm hoping this is a box within a box scenario so i can give this printer a proper unveiling ah there we go still kind of a nondescript box but if you can read the tape this is a cintra tech kit now this printer's actually been on the market since 2017 but there's only a couple videos out on youtube of it and i don't know a single person who owns this printer so why have we not seen more of these printers in the wild well the word affordable that i used in the introduction was definitely a stretch at six thousand dollars this kit finds itself in no man's land it's too expensive for the hobbyist and because it's a kit it's too much of a hassle for businesses where time is money and turnkey systems are preferred nevertheless i believe this kit serves a very important educational purpose because i'll be able to walk through the components that are found on not just this printer but also the high dollar sls printers that are revolutionizing the manufacturing process an increasing number of products are being fabricated through this technology from medical implants to automotive parts and the more people that understand how this technology works the better so this kit was purchased from a company based in miami florida they're known as 3d chimera and they did me right they included some extra tools to assemble everything a magnetic parts holder even though we already have our own as well as some surge protection equipment now sls is a little more complicated than the fuse fabrication process used by makerbots and creality 3d printers that have infiltrated the household over the years lasers mirrors and special powder spreaders are required for sls and don't worry we're going to go through all the components the increased complexity of this build is illustrated by the instruction manual being this thick did i mention that the instructions say that this will take four days to build from looking at the instructions i decided to divide this build up into four assembly sections first up we will build the enclosure next the powder handling components then the heating sources and finally the assembly that center tech refers to as the hat which contains the optics laser and brains behind the operation just to be clear this video is not meant to be a step-by-step guide for putting together this kit if you purchase the kit then you will receive a manual that will be far more detailed than i could ever be also i don't want to create content that would allow you to reverse engineer the centrotech kit without purchasing it because that would obviously hurt the company and while six thousand dollars is expensive it's still awesome that center tech released a kit at this price point because literally no other company is doing this so this video will be part time lapse to speed through the boring parts and part live commentary to talk about the important components let's start with building the enclosure the enclosure and most of the whole printer is constructed out of aluminum extrusion the aluminum extrusion frame will be filled with thick insulation because thermal management is critical for sls printing we will discuss this in more detail when we install the heating elements but during the printing process the powder is kept just below its melting point and the energy from the laser is just enough to make the powder melt if the temperature inside of the enclosure is off by even a couple of degrees and the powder will either fail to fuse or will fuse in unwanted areas depending on if the temperature is too low or too high respectively all right so we have the glass installed and now it's time to add some insulation again the insulation is going to be absolutely critical for making sure that there aren't temperature fluctuations within the chamber it'll also save us a lot of money because sls printers are are power hogs [Music] and with that we have a door it's a fancy door but it took a while i'm starting to think that that four day predictions probably uh spot on it's time to do the rest of the enclosure [Music] so right now we just kind of have a glorified box but one feature that i want to point out is that the inside of the chamber will get up to 150 degrees celsius and so to prevent the heat from leaving the chamber you know going through the aluminum extrusion to the outside world because you know obviously aluminum extrusion conducts heat really well there's actually these plastic washers that are on either side of the sheet metal so that's going to keep hopefully most of the heat inside the chamber and not connecting away to the aluminum extrusion those insulating washers are a simple thing to add but they should do a great job at keeping this chamber at a constant temperature now it's time to install the door and finish up the outside of the chamber [Music] [Music] this might be the most expensive box that i've ever built but hey at least it's really well insulated next we need to build the components that are going to go inside this enclosure to make it an sls printer now that's going to be a separate assembly so we can put the enclosure to the side for now [Music] here i've just finished up the skeleton for the powder beds as a reminder sls printing uses powder instead of filament or resin a laser melts the pattern in a defined geometry and then a new layer of powder is swept over the center layer with the sequence of events repeating until the print is finished this process requires two separate bins one acts as a powder reservoir and the other as the build volume where the part is being centered the powder is swept from the reservoir to the build volume after each layer is complete through the use of stages one in each bin and a spreader the size of these bins will define our maximum printable volume which is 110 by 110 by 110 millimeters yes this is small but that build volume is what is keeping this printer at this low of a price point larger sls printers are more difficult to uniformly heat spread powder across and scan with a laser so expensive provisions must be made fortunately the projects that i have planned will require small components printed by sls next up in the build i'm going to start working on the linear motion components for both the stages and the spreader the stages move in the z direction both up and down while the spreader sweeps from right to left the linear actuators that will control this movement will use lead screws and metal cables to move their carriages you won't see any timing belts in this machine because of the temperature of the chamber which maxes out at 150 degrees celsius it's too hot for neoprene the carriages will use v wheels that ride snugly down the middle of a piece of aluminum extrusion these are not the normal delrin or polycarbonate v wheels found on hobby's 3d printers and cnc routers because similar to neoprene the high temperature of the chamber would significantly affect their mechanical properties instead these wheels are made out of ptfe which has a surface temperature all the way up to 280 degrees celsius which is more than enough for this application it's really the need for a heated chamber that makes sls printing so much more complicated and expensive because it restricts the materials that can be used in the design [Music] [Music] so i just finished installing the spreader you can see it moves back and forth and i want to dive in a little bit more detail because transporting the powder from the reservoir to the build platform may seem like an easy problem but it's anything but that the quality of the final print is the sum of its layers if the powder is spread uniformly each time then robust and consistent parts can be printed if there are voids and clumps in the powder layer then your parts will suffer researchers have spent many years designing blades that deposit a thin layer of powder on the build volume some sls printers have even deployed more sophisticated powder spreaders such as counter rotating rollers which function as their name suggests the roller rotates in the opposite direction that is traversing it's been shown that this counter rotation stimulates the powder and increases its flow ability the point i'm trying to make is that the design of the spreader whether it is a blade or a counter rotating roller has been a serious design consideration for 3d printers for this reason it's kind of disappointing that the centrotech uses this ordinary piece of aluminum no special geometry nothing even worse on the edge that's going to spread the powder there was like a witness mark from the cutting process which i had a file out again it seems like another cost saving measure and this printer is designed for centering nylon powder and fortunately for us nylon powder has good flow ability and likes to be spread out into a uniform layer i bring this up not to put the design down i've actually been having a lot of fun building it but they do advertise this printer as an open materials platform and with that spreader any material that's less than ideal is not going to work very well but fortunately since we're building it it wouldn't be that hard to upgrade it with a different spreader we're now going to work on the back of the machine to install the actuators that are going to move the stages up and down [Music] [Music] so i just installed the stages within the two bins during the printing process the stages will move in opposite directions with the reservoir stage moving upward to supply more powder and the build volume stage moving downward to allow the part to grow in the z direction when i was working on the bins i was curious as to how the powder was going to be kept inside of them because the stages have to move and where there is movement there are gaps to plug those holes citrotec used what they are calling thermoseal fabric this fabric seals the space around the perimeter of the stages and the opening for where the stages connect outside to their actuators the fabric slides easily along the sheet metal and i really like the simplicity of the solution the instructions have the heating coil being installed before finishing off the actuators for the stages so let's go into a little bit more detail about heating with the cintratec sls printer there are three important temperatures the first is the temperature of the chamber which as i mentioned previously will be 150 degrees celsius when printing nylon the chamber will be heated by this large coil installed on the back of the machine and regulated by thermistors the second temperature is found at the surface of the build volume which will be just under the melting point of nylon nylon melts at approximately 178 degrees the surface temperature is regulated by infrared lamps this is a separate assembly that we will get into a little bit later the third and final temperature that the powder will experience is a temperature increase when hit by the laser which will be in excess of 180 degrees perhaps you're wondering why not simplify the design of this printer and just use a more powerful laser that can melt room temperature nylon well the rapid heating of many polymers will cause them to warp if you have experience with any other type of 3d printing you'll be all too familiar with how warping can ruin your final dimensions of your part also the uniform heating of the chamber will drive moisture out of the powder which improves the sintering process i will continue to talk about the heating process when i work on the lamps because it's just so critical for the success of sls prints [Music] after attaching the heating elements i went ahead and installed the limit switches for both stages now each stage has two limit switches one at the top of the travel and one at the bottom of the travel now these are read switches and they form a closed circuit when a magnet is brought into close proximity to them i have the top and bottom reed switches connected to my multimeter and my multimeter will beep when they are continuous when they are connected when they're closed so if i bring the stage up [Music] continuity and if we bring the stage to the bottom of its travel we have continuity there as well as you can tell from this demonstration the read switches are normally open for limit switches i prefer a normally closed configuration because with normally open you don't know if there's a problem with limit switch until you know it's too late you have a crash it's much safer to do normally closed however these reed switches are fully enclosed we don't worry about the powder getting inside of them and messing up the internals now there's also reed switches for the spreader both at the min and the max i'm going to tidy up these wires then we get to work on the lamp assembly which is going to keep that top layer of powder on the build volume hotter so it's easier to center so i got a little ahead of myself and forgot to add the overflow bin when the powder spreader moves the powder from the reservoir to the build volume excess powder will be scooped into this bin over here i also installed two m8 threaded rods which will translate the stages up and down while these threaded rods look to be in great shape i personally would have preferred a lead screw because of the higher precision and reduced wobble that comes with this superior design and really the price difference for a lead screw is budget dust when compared to the six thousand dollar asking price this example fits the overall build experience thus far the parts provided are sufficient for their purpose but if you're a hobbyist and saved up your hard earned cash for this kit you probably would be expecting components to be a little bit more premium you wouldn't think that you would have to deburr all the tab and slot components as well as the holes for the carriages just to get the bolts to fit through and the eccentric spacers to sit flush you know that there's going to be some post-processing when the reseller sends you a file with the kit now don't get me wrong you get a lot of stuff for your money and the optical components that we haven't started to work with can cost a pretty penny too if you do purchase this kit just make sure you have your expectations set accordingly now we can move on to the lamp assembly which includes three infrared lamps i'm not sure how many sls printers use infrared or ir lamps to heat the surface of their powder but i really like this idea with this setup we don't have to heat the entire chamber to as high of a temperature which is not only an energy saver but it also decreases the thermal history of the powder thermal history is a really important concept in sls printing and it's worth talking about in more detail when plastics like nylon are kept at high temperatures they experience both physical and mechanical changes examples of these changes include discoloration cracking and generally a loss of desirable mechanical properties plastics are made up of polymer chains and the thermal energy from high temperatures can cause these chains to break shorten or recombine in a process known as thermal degradation in the presence of oxygen this process is sped up the thermal history of a polymer is basically a record for how long the polymer was kept at elevated temperatures the longer the thermal history the more thermal degradation will occur with sls printing the thermal history of the part and powder will depend on the temperature of the chamber as well as the duration of the print if you reuse uncentered powder from a previous print then the subsequent part would have less favorable mechanical properties due to additional thermal degradation compared to the previous part that was made from fresh or virgin powder we will talk more about the consequences of reusing powder after we start printing for now it's just important to keep this concept of thermal history in the back of your head because with other printing technologies such as fff where the printing material is not kept near its melting temperature for long durations it's not something that you normally have to worry about the last assembly is what centrotec is referring to as the hat the hat will sit above the powder spreading assembly and it's fairly complex in its own right containing everything from the microcontroller to the optics but before we get to that assembly i want to talk about the laser the laser provides that last little bit of energy required to center the top layer of powder the size and shape of the laser beam largely determines the resolution of your prints in the x y plane a poorly tuned laser beam can also affect your z-axis resolution because if the energy density of the laser is too high then powder from previous layers can be centered adding unwanted material to the part on the other hand if the laser is too weak then your part would not have fusion between layers which could either result in print failure or a very weak part mechanically let's take a look at the light path of the laser now i'm not going to turn the laser on but i do want to stress that even a small laser such as this one can cause permanent eye damage if shined in the wrong direction also i should point out that my knowledge of optics is limited so please do additional research if you're interested in this topic and check out the comments below for corrections or more information this is a diode laser diode lasers are cheap compact and don't require as much power as other laser systems however diode lasers also have their disadvantages the big problem is that diode beams are elliptical the preferred beam shape is circular so that the resolution in the x-axis is the same as the y-axis this disadvantage can be overcome through these anamorphic prism pairs which transform elliptical beams into nearly circular beams by magnifying the elliptical beam in one dimension the circular beam that leaves these prisms will pass through a plano convex lens which is flat on one side and convex on the other that will focus the laser light the focal length should be around the distance from the lens to the first layer of powder because this is when the laser beam will be the tightest and have the highest energy density proper focusing of the laser beam is what allows even a small laser like this one to melt powder before the converging laser light hits the powder it first balances off two mere galvanometers which i will refer to as galvos for short when it comes to galvos there is a lot to talk about and this may warrant another video because these devices are also used in other 3d printing technology and their accuracy is pretty amazing but very fundamentally these galvos have a motor that precisely move a mirror the movement of the mirror will ultimately change location of the laser beam in the build volume when you use two mirrors you can precisely move the beam to any point on that first layer of powder in other cnc technologies you need a linear actuator in each dimension that you want to position your tool whether it is an extruder router or something else linear actuators need to be lubricated and assessed for wear and tear galvos on the other hand require very little maintenance if the mirrors are protected from dust also the gallons that come with the cintratec kit are closed loop which is a premium feature i have talked about in past videos in the context of servo motors but to summarize there is an internal feedback loop that ensures the mirror reaches the correct position after hitting both of the galvos the laser will pass through some glass to enter into the heated chamber before finally hitting the powder bed while the print volume is technically 110 millimeters in all three dimensions the center tech website lists a recommended build volume of 90 by 90 by 90. now because of math this means that the recommended build volume is almost 50 less than the total capacity why is this recommended build volume so much smaller now this is just a logical guess on my part but i believe center tech had to reduce the build volume because there's an important optical component that's missing the f theta lens the f data lens can focus a laser beam on a plane without this type of lens the laser spot size will increase as you get further away from the glass window where the laser beam is coming down from because the beam's distance is greater than the focal distance of the plano convex lens if you don't have an optics background like myself these concepts can seem foreign just understand that without an f data lens the beam will be in less and less focus the farther you get from the center of the build volume as the beam becomes less focused and grows larger the energy density will decrease larger beam means lower print resolution and possibly under centering which will affect mechanical properties it seems like the omission of an f data lens appears to be a cost saving measure that's definitely a bummer but again with a kit like this nearly any modification is possible i've got one more gripe the laser and anamorphic prism set is one of the few proprietary parts in this kit actually made by cintratec and it came soldered terribly there was continuity between the positive and negative leads so if i were to plug this in the laser would have been fried i had to shift the copper pad over and put in some hot glue to keep that short from occurring i'm all for saving money but on proprietary components that'd be hard for someone like me to replace the united states because this kit's coming from switzerland you really can't have shoddy workmanship [Music] [Music] [Music] so i just installed the steel cabling that's going to move the spreader back and forth remember we can't use timing belt for this application because the chamber is going to be too hot this printer uses three stepper motors the first two are for the build volume and reservoir stages respectively and the final one is for that pulley that's moving the steel cabling to translate the spreader back and forth [Music] oh [Music] [Music] believe it or not but by far the easiest part of the whole build was the wiring all the wires were of sufficient length and terminated with the appropriate connector let's go over all of the different electrical components that we have here there are three separate pcbs mounted in the hat the two that are standing vertically are the closed loop drivers for the galvos galvo drivers are a completely different beast than stepper drivers which many of us are familiar with the angle of the galvo's mirror becomes greater with the higher the absolute value of the voltage because the voltage can be positive or negative in turn the driver expects the microcontroller to send over a 10 volt differential range signal because galvos are so quick you need a fast microcontroller to keep up with it let's follow these silver braided cables that connect the galvos to the controller board this is a 32-bit sam microcontroller as the brain behind the operation remember this kit came out in 2017. that's two years before marlin the largest open source 3d printing firmware officially supported and popularized the faster 32-bit architecture there are some open source firmwares for running sls printers but they were not and in my opinion are still not refined enough for a commercial product so centrotech designed their own firmware back in 2017. if everything works without a hitch on the software side then that is a major pro towards buying this kit instead of diying one of the open source sls designs that have been popping up recently and with that i need to wire in the dc power to the microcontroller install some buttons fans hook up the mains and drop everything into the enclosure if i'm sounding excited that's because we're so close so uh just stick with me a little bit longer and we're gonna get this thing printing [Music] so [Music] so wiring the mains was super easy because sensortech provided this awesome wire harness which provided power to both of the dc power supplies as well as to the microcontroller now the microcontroller takes ac power for the heating elements and dc power for all the digital logic and the stepper motors and galvos now before i attach the top of the hat which contains fans to keep the electronics cool and before i drop this entire assembly into the enclosure we need to go through the initial commissioning the initial commissioning is the process of making sure that everything is working correctly the stepper motors are rotating in the right direction the heating element is turning on the thermistors can read temperature differences when the heating element is on versus off the galvos need to be calibrated so that they're scanning the laser within the build volume there's a lot of things that we need to do to make sure all the components are working together so that this sls experience is flawless after turning the machine on for the first time i was surprised to find that the spreader and the stages started to home automatically that process completed successfully so it looks like we installed the motors and all the linear motion correctly so next i downloaded the software which is centra tech central to start the laser calibration process now this was a little bit difficult to film because small adjustments made a huge difference but here are the basic steps first you want to make sure that the laser hits the x galvo right in the center now to get it in the center from left to right i actually moved the entire laser module and to get it in the center from top to bottom i moved the galvo assembly after it was centered the laser beam hits the y galvo and then goes down to the build platform as we talked about previously and it lands on this calibration pattern now we need to make sure that what is being generated by the galvos lines up with this pattern and so i actually moved the galvos in their housing so that they matched uh that pattern finally i had to scale the area that the galvos were scanning in both the x and y dimensions so that they matched the build platform and with that it's time to drop this whole assembly into the enclosure with this thing fully assembled we're so close to printing the last thing i want to talk about is calibrating the laser beam spot size the diameter of the laser beam that comes into contact with the powder bed is directly related to the x y resolution of the printer now because the laser energy is constant the spot size will also control how much heat is pumped into the powder with smaller spot sizes resulting in a higher energy density the way to adjust the spot size is to move the focal lens forward or backwards citrotec recommends moving the lens as close as possible to the galvez to yield a 200 micron spot size for highest resolution and energy density which is what i've done the final components of the assembly include the top of the hat which has four fans to keep the electronics cool the electronics and optics are literally sitting above an oven so airflow into and out of the hat will be really important we need to fill up the printer with powder but first let me talk a little bit more about nylon nylon is a synthetic thermoplastic polymer that belongs to the family of polyamides or pa for short there are a lot of different nylons but i have pa12 here pa12 is stable against impact chemicals heat uv light and water making it ideal for both rapid prototyping and production really nylon is the key to sls 3d printing's functionality and versatility and will allow me to produce parts i don't have to worry about degrading if left in the sunlight or in a humid environment for too long if i open up this jar and pour a little bit of powder on my workbench you may be able to tell this powder has high flow ability it's almost like pouring water the individual particles which are approximately 60 microns in size don't stick or clump together i want to point out that whenever i handle this powder i wear a respirator and gloves nylon dust is a minor respiratory irritant so the mask is a protective measure in theory the 60 micron particles shouldn't become airborne but it's very apparent when you pour the powder that there is much finer particles that get into the air the gloves are to protect the powder oils from my hands can contaminate the particles causing problems during the sintering process the last safety precaution i take is to be in a well ventilated room when printing i'm in my garage so i just crack the garage door but there is a noticeable smell in the air when the powder is heating up so i wouldn't recommend running an sls printer in your living space let's go ahead and fill the reservoir with just enough powder to perform a calibration print to check the dimensional accuracy of the printer we don't want to overfill the bin because that would expose unneeded powder to thermal degradation which is wasteful inside of center tech central is the calibration model that we need with some back of the envelope calculations we can determine how much powder we need this measurement is going to be height how far down does the reservoir stage need to be this part has 50 layers and the layer height is set to 100 microns so we need five millimeters of powder to complete this print however we need a little extra powder to prepare the build platform with a nice flat starting layer and some top layers to insulate the part during the cooling down process i add an extra 30 millimeters to the final height because really the worst scenario is to run out of powder this means i need to move the reservoir stage down 35 millimeters and fill it to the brim that takes care of the reservoir to prime the build platform i add a little powder before having the spreader sweep across the build volume and i fill in any holes and gaps with a little extra powder and that's it i don't have to adjust any parameters because i'm using cintratex validated powder and the pre-loaded calibration model i can now click start print and wow that's going to take an hour and 45 minutes to heat up now the chamber can quickly get to that temperature but the powder is slower to heat up it's very important that the powder is at a uniform temperature in this case 140 degrees celsius so that there are no hot spots that will spontaneously melt and no cold spots that won't center when hit with a laser the heated chamber also bakes out any water in the nylon nylon is hydroscopic meaning it is able to absorb a lot of water from the air under normal conditions this is especially problematic when printing nylon as a filament through an fff 3d printer but that won't be a problem with this heated chamber on the sls printer sls is definitely slow to get going from preparing the powder bed to waiting for the heat up process but our patients should be rewarded quick clarification i had said previously that the chamber would be at 150 c for nylon but i later read on the centrotech website that keeping the chamber at 140c increases the life of some of the internal components i'm not sure what components would benefit from that 10c difference but for now i'll just follow along with this preset profile of course the chamber temperature and many other settings like the laser speed can be changed this is an open platform that will work for powders outside of the centertech ecosystem which is really great it's going to be a little hard to film the printing process because all we have is this little window but this is why i thought this would be a great project to build on camera because you've already seen all of the inner workings of the printer and can visualize how these components are working together even with this limited view [Music] so morning so man look at that laser raster across the platform it may look like there are multiple laser beams but that is just your eyes playing tricks on you those galvos are moving the beam so quickly it just looks like there's multiple lasers working together notice how the powder isn't melting into a puddle but instead looks leathery this is what we want because a puddle can't be controlled the powder particles within the laser spot are melting very briefly and then cooling down into this meta-stable state there's actually a lot of science behind controlling the polymer melt but just know this whole process has been optimized to be very precise selective laser sintering is a bit of a misnomer because centering is when two particles fuse below their melting point and in most sls printers there is a spectrum of both centering and melting of the powder if you look really closely then you can see a small line forming in the middle of the layer that is from that mark in the spreader that i guess i wasn't completely able to get rid of at a later date i will make a new spreader on my mill now you may see the heat lamps flickering they are changing intensity to bring the surface temperature of the powder to 170c just the right temperature for the laser to center at this process will continue for another 40 layers or so the part will then be encased in a couple layers of hot powder before the chamber heating turns off this allows the part to cool down more slowly and evenly to prevent warping in some ways sls is slow because you have to wait for the chamber to heat up and cool down but did you see how quickly the laser skin across the surface printing only one part that only takes up a small amount of the cross-sectional build area is not the correct approach you want to print as many parts as possible in the build volume we will talk about nesting parts in the build volume before our next print but this print is finally done so let's see how it did we can raise the build volume up to get easier access to the part in the surrounding powder i haven't finished my powder collection system yet so i'm just scooping the extra powder out with a little measuring spoon this whole post-processing powder removal step will be optimized in the future i've seen people using bead blasters and tumblers to quickly remove excess powder i might stick with a vacuum and a brush for cost saving reasons wow the texture on this part is really awesome it's almost impossible to discern the layers just from a visual standpoint you might be able to tell why a part printed off an sls printer is very marketable let's check the dimensional accuracy of our calibration print it should be 40 millimeters in the x and 40 millimeters and the y that is pretty close for a first print and air appears to be about the same in both dimensions the z-axis is even closer to its true value of 5 millimeters inside the software we can compensate for these dimensional inaccuracies by applying small scaling factors now i would reprint this calibration part after making those digital adjustments but it's kind of a waste of powder so i'm going to start printing models that better showcase this technology and i can take dimensions off those components to check for accuracy speaking of how to waste let's talk about recyclability unless you are printing a cube the size the build volume you will always have some powder left over that has been heated for as long as your previous print sls powder is expensive at 160 dollars per kilo so being able to use this leftover powder again would be great for the wallet unfortunately you can't just stick this powder back into the reservoir and start another print as we discussed previously this powder has undergone some amount of thermal degradation and as a result does not want to flow as well as fresh powder less flow ability means that it's harder to get that uniform layer of powder also use powder is more likely to aggregate and needs to be passed through a sieve to remove chunks of powder now that you've seen the printing process just think about how a big chunk of powder would tear through a thin uniform layer of powder messing up our layer height i ordered a 200 micron sieve which should get rid of most the chunks and hopefully won't clog too easily but that hasn't come in yet so i will just save this used powder for later and i'm going to continue with fresh powder as i mentioned the most cost effective way to run a sls printer is to fill the build volume with multiple parts through a process known as nesting while build orientation within the build volume is extremely flexible with sls because the powder from the previous layer will support all the overhangs there are still several guidelines you should follow to minimize print failure what does print failure look like on an sls printer typically it occurs when the part curls off the build volume and gets hit by the powder spreader here i was printing a lattice structure which was just outside that 90 by 90 by 90 recommended build volume one of the corners must have been in a cold spot because the laser energy caused that corn to warp and lift off the powder spreader quickly made contact with it dragging it across the build area scarring the adjacent part at that point you pretty much have to end the print this leads us to the first rule of thumb for nesting keep your parts as close to the center as possible this is where the powder temperature will be the most uniform and the laser spot size will be the tightest the next recommendation is to position parts in order to evenly spread out the thermal energy through the build volume take this pyramid for an example the tendency may be to place the pyramid upright however in this orientation the laser is going to have to spend much longer scanning the bottom layers of the pyramid compared to the top layers those bottom layers run the risk of off target centering of powder beneath them because you're pumping so much energy into them also that excess energy could cause warping and curling of those layers if i click the optimal part orientation button the software flips the pyramid sideways now the longest plane is in the z-axis which is better but we still have quite a big surface area here in the middle i would probably print this pyramid at a 45 degree angle the smart orientation button only takes into consideration the model that is highlighted so it's not very useful when you're printing multiple parts let's say i wanted to print a group of six army men when printed at the same starting height their bases form a large cross-sectional area a better way to set up this nested print would be to flip over every other army man or move them up and down into different planes just remember that the slowest part of the print is spreading a new layer and heating it with the ir lamps so i guess that leaves us with two tips the first is to minimize cross-sectional area of centered material in each layer and the second is to minimize the number of layers to decrease print times those two tips can be at odds with each other because usually decreasing one increases the other that was just a quick introduction to setting up prints for sls i want to share this different approach to arranging models because it's not as intuitive as placing the largest flat surface of a part on the build platform like you would do with fff or other additive manufacturing technologies sls is very unique in the way that the powder supports the growing part so i'm over here at my plasma table and i've been itching to make something useful with this printer and i have just the part to mount my hcp plasma torch to my cnc table i've been using a pla holder however the humidity from the water table as well as the uv radiation from the plasma has made this holder extremely brittle it's even cracking an engineered plastic like nylon would be so much better for this application so i'm going to print a new torch mount this holder will have a couple of tricks up its sleeve that i will walk you through after the print this new torch holder design has seven parts that i will nest and print at the same time i snuck in a benchy as a fun keepsake because i had a little bit of room i have the parts as close to the center as possible and all are within that second wireframe which is that 90 by 90 by 90 safe volume i've already prepared the powder bed so we can hit print this print is going to take about 12 hours so in the meantime i went ahead and put together this powder extraction system it's basically just a cyclone separator that is attached to my vacuum the separator will remove the nylon powder from the air before it has the chance to clog the filter in my vacuum the powder will collect in this basin which is a great way to store it and prevent it from getting contaminated now the process of sucking up the nylon particles to a long plastic hose generates a lot of static electricity without a way to safely discharge this energy i would get zapped if i touch the frame of the printer because it's grounded not only does this hurt but it also has the potential to damage electrical components in the printer so i wrapped bare copper wire around the hose and grounded it this has worked wonders and i haven't been shocked yet [Music] so [Music] so the print is finished and i gave it about two hours to cool down so that the part cools homogeneously and we get really nice mechanical properties throughout the entire part [Music] after sucking up that excess material i will transfer the powder cake which is the part and all the powder still stuck to it to a secondary bin where i will use a nylon brush and these metal tools to get the rest of the powder off the part this process is very similar to an archaeology dig and it's like you're rediscovering your part which is a lot of fun i finished the de-powdering process with a blast of compressed air man the text on the torch mounds face is super crisp as you can read it says smart on the top has a minimalistic dr d flow logo in the middle followed by the words mag on the bottom with a little picture of a magnet this text is extruded one millimeter from the surface of the part so i hope you can start to appreciate the resolution of this printer we'll talk more about what this text means a little bit later these two components have holes on almost all of their faces which i'm going to use to insert some threaded inserts with the use of my soldering iron so when i modeled these holes i put a slight taper on them so the heat inserts go in straight so unfortunately these are m4 inserts and i don't have an m4 tool for my soldering iron so i'm going to use one size smaller it'll still work fine [Music] this part bolts onto the floating head assembly of the plasma table if you're new to my channel then check out my plasma cutting playlist to learn about the floating head and other cnc plasma concepts before i attach this piece i need to insert these two ball bearings which will locate the components that actually grab the torch to this plate this will make more sense in a second now the ball bearings have an m4 threaded hole on one of their sides which allows me to attach them to this piece there are two holes on either side of this component for magnets i just slide them in and then i'll cap them off with these two end pieces now this end piece needs two threaded inserts these are m3 though quickly insert those before i forget here's the part that will actually grab on to the plasma torch there are two domes located on its backside that will locate to these ball bearings there are two more holes on the top and bottom for magnets and that's what keeps this piece attached to that one magnetic torch mount we will cap both ends as well to keep those magnets in there and now these two pieces should stick together which they do and because those ball bearings the components don't rotate it's going to hold the torch straight it takes about 30 pounds of force to separate these two components which should be strong enough that i can still touch the torch off materials and use that initial height sensing but weak enough that if the torch side swipes apart that it'll be able to break off without the torch breaking so that takes care of the magnet part of this assembly now the smart part comes from incorporating this limit switch on the back of the component that mounts the floating head this just bolts on up here so the piece that grabs onto the torch when it comes together with the piece that mounts the floating head it presses the limit switch button if the torch were to fall off the limit switch would be untriggered which would alert the microcontroller to halt all motion and cut power to the torch which will be a great safety feature so that's kind of why i included that little smart word in the logo so it's a little bit of a bulky limit switch but what's cool about it is that it's waterproof and super rugged so it's gonna be great for the plasma cutting application the last part of the assembly is to attach these two cables this will keep the torch mount part from falling too far away from the floating head depending on your plasma table setup you know this could be important if you don't want your torch to drop too far which that could break it as well in my application not as important but just going to throw those in there [Music] [Applause] [Music] so here's the final assembly with the torch installed it looks really awesome the fit's great the rougher surface of the parts does a great job of gripping the smooth surface of the torch which is awesome now i'm going to give a quick demonstration of the magnetic release over on my plasma table but i don't want to purposely run my torch into pieces of metal so i'm going to use this dummy 3d printed torch look alike for this demonstration purposes [Music] that worked like a charm as soon as the torch sideswiped the metal plate it popped off and the microcontroller sensed a change in the limit switches state so it stopped moving now you couldn't tell with this demonstration but the way i have everything wired when the torch falls off the trigger would be cut ceasing the plasma stream the reason i went into so much detail about making this torch mount because these functional applications are where sls printers excel in my opinion until the cost of nylon powder comes down sls printers are going to be reserved for commercial applications perhaps such an application is producing small quantities of specialty parts in your garage to sell online and with leftover powder you can print some different gadgets and hobbyist parts let's just say i'm seriously considering this route with either the magnetic torch mount or one of the many other niche ideas that i have now you may be wondering why x is here and that's because i was originally inspired by this bmg extruder that i installed on this printer for my how to build a 3d printer video the housing for this extrusion drive is actually printed on an sls printer and it has held up remarkably well for almost a year and a half of use this is a great example of how a successful business has incorporated sls parts in their product line for our last print i want to create some models that showcases the benefits of powder bed fusion because with this printing technology the powder as we've talked about before supports floating structures so with this type of printer you can create joints that will be movable as soon as you pull them out of the powder bed which is really cool i'm also going to print a torture test in the form of a lattice structure just so you can get a feel about the resolution that this printer possesses however i'm down to my last kilogram of fresh powder and i want to make it last a little bit longer so i can actually take some used or aged powder and add it into this mix and perfect timing my sieves arrived so we're going to take powder from the cyclone pass it through these sieves to get rid of any chunks that would otherwise affect the print quality and we should be good to go so i only use my cyclone separator to suck up powder from inside and around the printer so really we shouldn't have to worry about contamination i'm going to go ahead and put my mask on and we'll see how much powder i've got built up in here not too much so just simply transferring the powder from the big container to the little container has already made a mess and let me tell you i now understand why companies like formlabs sell fully enclosed and automated systems for sieving and mixing powder you know it's not my budget but if you have the money to splurge on such a system i think it's going to be well worth the money just to avoid the mess so i'm now going to take this age powder and pass it through two sieves the first one is a 500 micron sieve which will get rid of all the chunks and the second sieve is a 200 212 micron sieve which should hopefully leave just a nice fine powder now technically the powder is as small as 60 microns but if i use a really small sieve it would take forever to go through all the powder hmm so i separated the two sieves on the left is the 500 micron sieve with the larger pore size it has all the really big almost pebble like powder that's you know kind of loosely fused together this powder is probably really close to where that part was being printed and on the right we have the 212 micron sieve much smaller particles but still these particles are not as fine as the sieve powder which is very similar to the fresh powder but there are some subtle differences in the mechanical properties of this age powder that's been exposed to temperatures for a long time so we can't even though we have a lot of it we can't just throw this into the chamber we're gonna have to mix it with the virgin or fresh powder so the way i mix the powder is through this graduated cylinder now center tech recommends mixing old and new powder at a ratio of three to seven that's three parts old or age powder and seven parts new powder i just quickly not too precisely measured out in this graduated cylinder and i pour it into a new container and label it mixed oh so that's 300 ml of the old powder now here comes the fresh powder now obviously you don't want to spill this on the table because that kind of defeats the purpose of making it last longer by mixing it with uh old powder [Music] all right so we'll take two graduate cylinders to fill up one of these containers now that i got everything cleaned up we will throw the mixed powder into the powder reservoir and prepare the build surface [Music] in addition to a little ball and socket joint and the lattice i'm printing some fun widgets including a two-way screw and two bearings i also threw a spring in there to show off the mechanical properties of nylon which might surprise you [Music] [Music] [Music] so [Music] the ball and joint socket turned out nicely i had modeled this infusion for there to be 250 microns of radial space between the ball and the wall of the socket i can get some really smooth motion out of it and because i didn't have to use slits to allow the ball to pop into the socket as you would have to do if these parts are printed separately the joint is much stronger so the lattice structure was an interesting one the wall thickness of the lattice is one millimeter so the printer had no problem resolution wise however it was very difficult to remove such a small and complex structure from the powder cake without breaking it i initially clipped the side of it with my vacuum and that's why there's some damage to it so my conclusion from this print is that while the printer's theoretical resolution is around 200 microns which is the width of the laser beam features will have to be quite a bit larger and as a result stronger in order to remove the excess powder from the calibration print and the torch mount you wouldn't think that nylon has much flexibility but check out this spring nylon is more elastic than pla and abs this property is what allows nylon to be so tough it can absorb tremendous amounts of energy without fracturing i took the top of my torch mount and hit it with a hammer now that's not a mallet that is a dead blow hammer which i hit pretty hard and not a single scratch a letter didn't fly off it's it's incredible the bearings weren't as satisfying i pulled these models off of the internet and they are meant to be printed with an fff printer and there was just way too much space between the bearings which allowed them to get stuck after rotating them a couple of times this two-way screw is just satisfying to play with there are two nuts one that screws on clockwise and the other that screws on counterclockwise of just a fidget toy but it's a lot of fun i forgot to show you guys benchy from the torch mount print this turned out pretty awesome too now one of the main reasons benchy's used is to look for stringing and see how well you can bridge gaps and you know neither one of those are factors when it comes to sls printing so this model looks great there's probably some viewers wondering why you would spend the money time and put up with the mess of an sls 3d printer if you can already print nylon filament on an ff printer i've already mentioned that nylon filament is hard to print via fff because when it sits at room temperature it's absorbing water out of the air resulting in foamy or weak extrusion in sls the heated chamber bakes out any absorbed water before the print and during the print the temperature is sitting above the boiling point of water so the nylon powder is extremely dry but in general nylon filament is kind of finicky to print with its slight elasticity can be difficult for the extrusion drive of an fff printer to attract especially if you're running a bowden setup with sls printing there is no retraction separate features are created by turning the laser on and off now another benefit of sls over fff is that you get much better fusion between layers and we take a closer look at benshi's hall you can barely see the layer lines fff printed objects tend to be strongest in the x and y axes and weakest in the z axis direction due to the strong bonds between layers sls parts have more isotropic mechanical properties and thus they're strong in all directions finally the last major difference when printing nylon on an sls printer versus an fff is the print time the print duration of an fff printer scales with the volume of the part generally a print that is twice as big will take twice as long this is not the case with sls because the laser is able to scan across the powder bed at such high speeds the powder spreading action is the only time consuming part therefore printing only scales with part height when parts are properly organized in the build volume to minimize their z height an sls printer can produce parts very quickly i don't want to be too rosy because it seems like for each of the benefits of printing with powder there's also a drawback perhaps the two biggest drawbacks are the mess and the small build volume these problems can be solved by purchasing very expensive printers with large print beds and powder removal stations but those are not really options for those on a budget so i'm going to focus in on the one complaint i have with this hobbyist class printer after completing a couple of prints i've found the non-uniform powder temperature during the printing process to be a problem you can't print parts near the edge because these regions are either too hot or cold and as we saw earlier there's a strong likelihood that the part will curl up and hit the powder spreader because powder still has to occupy these edges there's always a lot of wasted powder after each print if you're trying to make a product with this printer then your material costs are going to be about twice that of what they should be because of this powder waste i also noticed that some of my prints even when printed within the safe zone came out slightly warped this occurred even after allowing the printer to properly heat up and cool down while using the prescribed printing parameters for this powder but it only happened every now and then which makes it difficult to troubleshoot this led me to the conclusion that this was also probably a result of non-uniform powder heating now i expect this would be a problem for all diy and budget sls kits because the thermal control of the powder is very complicated and will require multiple heating elements and thermistors throughout the chamber to monitor for temperature fluctuations which would be expensive and difficult to build outside of a factory setting to end this video on a good note let's quickly summarize what we accomplished here we built an sls printer in my garage that is capable of producing absolutely stunning parts out of a performance plastic this is accomplished through a set of optics that can scan a laser beam across a layer of powder in a powder handling system that can sweep thin layers of powder over previous layers pretty spectacular and i'm so excited with the way it turned out if you enjoyed this additive manufacturing content then definitely get subscribed because this printer is probably not even going to be the craziest thing i build this year i have enough linear rail to build a roller coaster and enough motors to power a small village a lot of exciting content coming up so i'll catch you guys in the next one

Channel: Dr. D-Flo

Views: 184,560

Rating: 4.8770833 out of 5

Keywords: Powder Bed Fusion, Selective Laser Melting, Laser Printing

Id: OjjczQkK2zM

Length: 76min 4sec (4564 seconds)

Published: Sat Mar 06 2021