Vision Miner’s Post

Transcript

Alright guys, we're here at Amog 2024 and Aon 3D brought their high lows. We're gonna jump right in for another deep dive look with Nick Parker, one of the head engineers on the project. Nick, how's it going today? Hey, good to see you man. So super happy you guys were able to bring this and show it off. Yeah man. But really, what it comes down to it, what we wanna know, what we need to know is why would you get something like this over the previous industry standards of something like the Fortis or something like these big high tech machines. What makes the high low unique and different from all these other high temp chamber machines? Yeah, why would you buy Hilo? I mean, it depends who you are. If you're just trying to make jigs fixtures, you're chasing reasonably easy printing in high performance. Materials that are gonna be able to meet your use case. This is the first open market high temp machine that you're actually gonna like. It's not gonna drive you crazy. It's more or less just gonna work at the touch of a button. We've got a thermal simulation. Even if you're trying to do all you're trying to do, high performance materials that are usually going to be a very high skill thing. You know, we hire a lot of PhD, it'll mostly just work. And then at the other end of the scale, if you've got genuine cutting edge problems, if you're trying to certify parts for flights, if you've got novel plastics, if you're trying to do the crazy stuff. And you, you know, you consider yourself a leader. This thing's got every sensor you could possibly want. It's crazy repeatable, super deterministic, and it's pretty much the best possible Rd. platform. Nice. Yeah, we definitely can relate to the, you know, the high temp materials are generally very, very challenging on open systems you're dealing with. You've got temperatures, but all those geometries and everything are going to be different and behave differently. So a lot of it's on the software side. Can you tell me a little about how you make this easier with all the new software and AI and whatnot behind it? Yeah, sure. So a lot of the easiness is just, hey, there's nice work clothes on here. You can see some of the buttons. They play nice. Yeah. But where we really shine is we've got a thermal simulation that's actually running through your parts about a quarter second at a time, about 1/4 millimeter at a time. And it's pretending to deposit every tiny little Fleck of plastic and then simulating over the full print duration. Hey, how hot is each of those little drops going to be? So that matters because when you come back and put another bead on top, if the previous layers got them too cold, it's going to be. Beautiful. It'll look like a nice part, but it will fail at right half your expected strength, right? And then also, if you're gonna overheat, we'll catch that too. But overheating doesn't scare me that much. You can see. Worst case, you waste some plastic overheating right? Under heating, you're gonna like, crash a plane, right? And then on the other side of it, once you've got a beautiful plan from our thermal simulation, this machine will do a beautiful job executing it. And it'll tell you if it manages to mess up, right? So if you've got a tiny flecks of plastic in there, if you gave us some junk material that's undersized and the diameter sensor somehow misses it, if anything. Goes wrong, those sensors are going to light it up like a gunshot wound. So one of the big things is the the not only the simulation. So you know where the part might mess up based on machine learning data from thousands and thousands of prints, but. During the process, it's using all the sensors to record what happens and what happened, so you can see how hot it actually was. One of the biggest things we see on the open systems is like Ultem 10/10 will BLOB up on the nozzle and it'll deposit that BLOB somewhere on the part and then it's jumping it. But if you're gone, you let it print overnight, you might not even see that. Might be internal, maybe in the infill. So with the full sensor gamut throughout the process that's recording, you can actually detect that. Yeah. Why is that important? And who's that important for? Yeah, it's important for I mean really anybody. If you're trying to learn how to run the machine. Yeah. This is sort of the benefit of four sensing. The people don't think about. We're all very good at 3D printing. We know how to look at this chunk of plastic and say Ohh exactly. XY and Z went wrong with that. I know how to fix that. Off I go back to the slicer. Easy peasy. If you just bought a new machine you just joined the company. You're growing fast. It's beautiful. Hilo is treating you very well. You've bought more. You're poor. Operator does not know how to read the tea leaves in a chunk of Alta. And so the four sensor data says, oh hey, look your first. There's a little bit over extruded, like you might have let that hydrate a little bit. So it really cuts that learning curve. Yeah. But then where it really, really shines is like I said earlier, if you're trying to do certified aerospace, if trying to do anything with any criticality, if you care if your parts are strong, then you want 4 sensors on there to tell you if they come came out right. So what, what type of anomalies during the print and defects can you detect with this tech? Yeah, so there's a lot of boring ones. We've got sensors down in the feet, assists in the material management. Four. OK. It'll tell you, hey, are you giving it enough plastic? It'll measure the diameter of everything coming through and compensate that. There's tension, compression sensors, but what we're really excited about is force monitoring and the simulation. So the two scariest things, if you've got a critical use case, are cold welds and debris inclusion cold welds. You have a part that looks totally fine, comes off the machine maybe prettier than your average parts, but because some section of it got a little bit too cold before that next layer came by, it's just crazy week. We're talking half of your rated strength. You put that into use, it's going to break. And you know, maybe you just printed on a cold Tuesday, so most of your parts are fine. You're in a serial production context, you're going to get bits. And then the other scary one is debris inclusion. So you have a little bit of something to depositing on the nozzle. You asked it to do some crazy sports structures didn't quite work out. The parts kind of fine. So you know, comes out look and sensible, but you don't know deep inside of that part in your most important load path, there is a big chunk of trash, big old air voids around that chunk of trash. And again, you're going to fail at half your rating strength. Yeah, that's one of the biggest things, you know. But he's like, Ohh, yeah, 3D prints aren't strong. And the Z axis adhesion, it's like, well, this massively increases this. Yeah, well with the simulations and everything, how close to, well, not isotropic strength, but how close to the full strength in the Z layers are we achieving? Depends what material 70% and consistency, yeah, that's where I was about to go is consistently right. Because I mean, if you're a competent engineer, you can say, OK, cool, I've got this strength and X&Y, that strength -.3 and Z and you can design around that. That's all fine. The reason people don't fly FDM parts, The reason people don't trust FDM parts is because the error bounds are way the hell too big. And so that's what I was saying earlier with determinism. This thing's got a ton of work put into it to make sure that you've got super accurate positioning. Even at 250 Celsius, your bed still dead flat. Your bed mesh is still absolutely beautiful. So that you can say, OK, hey, if I make fifty of these on different high lows on different bases around the world, they're all gonna be within some reasonable percentage of each other and I can have a safety factor that's, you know, a normal safety factor, not like 2 digits. Should take a quick look inside anything you want to show. Chamber flow is pretty obvious. Everybody that knows what they're doing and does a bit of a Whirlpool setup. So you've got outlet inlet on one side, inlet outlet on the other. That gives you this nice turbulent mixture in there, which turbulence sounds bad, but it lets you stay uniform. As you print different parts, if you try to do a perfect laminar thing, you've got beautiful flow for one tensile coupon. But you put anything real in there and you're gonna have hot spots and cold spots. You're not gonna know what's happening. It's gonna be super sensitive to tiny deviations. If we look in the back here, we've got something pretty proud. Proud of. This is the Purge inspection board. So that clicks out to a list removal. That's brushes that the machine can come and wipe off on. You can purge material, drop it in here with a little flicker. And the drawer goes back in. That flicks open your appellative purge material, falls down a pipe to the back of the material drawer. That sounds very simple, but notice this is outside the heated environment, almost every other high temp system. We'll just give you a bucket in the oven. The bucket works great. One day you forget to empty the bucket. Now you need a new bucket. There'll be 24K cameras, one in each in production. The other third of that drawer is another camera. That's a microscope bounced off of a mirror. So the microscope hides in the back, stays nice and safe even at 250 C and that. Let's you get, I wanna say 6 Micron pixel pics. Pixel pitch. Yeah, images inside at the tip of the nozzle. So I don't know if you've ever done XY offset. Well you have. I don't know if you, the viewer, have ever done XY offsets on a high temperature machine. It is not fun. You print some nonsense, Takes like an hour. On this machine it's about two minutes. Each nozzle just comes over, gets a little selfie taken. The software compares the two of them and off you go. Yeah, yeah, simple solutions that are highly effective. Like I said at the beginning, if you're just trying to get good. Arts in high temp materials. You will actually like this machine, right? And I don't think anything else on the open market can say that. So it's kind of like this eliminates a lot of the OK, we're gonna print it 16 times until we figure out the exact way, and then we can do it for the most part consistently. This is like, eliminates a lot of that guessing and checking with the AI, machine learning and just with the general workflow of it, that's awesome. What kind of build plates are you using? This one CFP, We really love it when we can get away with it. It's carbon fibers, magical. It shrinks. When it gets hot, it's really, really likes to let go of your part. When it cools off, Yeah, super easy. Some materials don't like that. So this is a vacuum Chuck. If you need to put something special on there for the pickier plastics, super easy. All right, So one of the huge things in the industry right now I'm gonna shut this is print speeds. You see people taking any kind of, you know, core XY machine, they're like, look, I'm going 1500mm a second and then you've got some companies that went super fast, but it physically doesn't necessarily make sense. You're flinging material everywhere. How fast can you reliably print on this machine? Hmm. And depending on the material, of course. Depending on material. OK, but so this little guy is a high, low hot end. If you're watching this on YouTube, you probably know vez 3D. And the guts of this thing are very, very similar to a Goliath, except integrated for high temperature. So this outer tube here is invar. That's the structural system that allows it to be dead stable as you heat it up. So the heater core in there gets one 1.2mm longer when you send it to. 50C The invar outer tube doesn't change at all, so your nozzle stays in the same place on the tip. Here you have a nonstick shield with tiny little holes in it for compressed air part cooling. Part cooling isn't super often something that you use in high temp land, but if you've got a thermal simulation then you can say, all right, I'm going to hug the hot side of my processing window. I am never, ever, ever going to have a joint that gets too cold. And for the four or five spots that the simulation tells me you're going to get too hot, I'm just going to blast them there. You were asking about print speeds when I pulled this thing out. So 100 cubic millimeters a second in PET G and other commodity materials, 3040 and Altum. Depends on your bravery for peak. Yeah, I like that answer because it's cubic cubic millimeter per second. Yeah, right. As opposed to the gantry, 100mm a second. The gantry will rip 600 at 1G. But that's, you know, you put a junkie nozzle on that. That's not gonna do anything, right. Speaking of non junkie nozzles, that's tungsten carbide. That means very good thermals, very good heat transfer, and extremely good reconnaissance. And never gonna wear out. Yeah, exactly. So you can run all the carbon fiber composites, glass fibers, nanotubes, whatever you're running through this machine, you can pretty much do it. Another cute detail on here. These little guys right here. So it takes 2 screws to secure this one that goes in that groove, one that secures a pair of heat pipes onto this side of the heat sink. So there's no water cooling in this system, and we're not dumping gallons and gallons of compressed air to keep the cold side of your heat rate cold. All we do is. They have effectively a CPU heat sink reaching up into the cold plenum here. Normal fan works like 1,000,000 bucks. They were printing ULTEM 9085 live at the show. That's something you don't always see. So and that was successful. Do we have that part here? Let me actually, yeah. So actually this is a good segue. This part got a bit by the trade show curse. So there's of course great big chunk of debris in here. Very unfortunate. However, it's an accidental demo because over here you've got the force data and let's check this out. Shot wound where you have the debris inclusion. So yeah, literally right here we can see that defect that's in the part. So all that debris in there and that's what you'll see on a lot of machines and you can see it right in there on the actual. Recording of data, Yeah. What do you call 204 sensor? Yeah, 224 sensor, sensor history MAP. So I mean in action, yes, it failed on the show floor. It had an issue where this would not be flight ready and it detected it. And that's just, I mean that's a perfect example of, hey look, this works, you'll know. Whereas that could have been hidden and he'd throw this up somewhere in it and you were asking another stuff to brag. There's one thing my marketing guy doesn't think is cool. The engineers on YouTube will alright, if you look at this machine, the entire structural loop, minus the heated bed itself is cold. So that's called a a thermal structural loop. And the point of that is that this thing gets to 250 Celsius in about an hour. On any other high temperature, you're going to wait about two more hours after that for the structure to finish thermally expanding. Here you just press print soon. Not off you go. That is really great. You know, heat soaking is one of those things where a new person go in, they turn it on. Hey, I started a printing and it's all messed up. It's like, did you let it heat up? Well, how long do I get? Well, which machine is it? How big is? How much metal are you really soaking? How cold is it in the shop, you know, versus this thing self-contained. What's that little noise? Right there is the purge from the compressed air dryer. So this drawer down here stays at a dew point of -25 at all times. If you put just about any polymer in there, it'll stay dry indefinitely. If you put low temp stuff in there, it'll get dry after a couple of days. That little hiss you here right now is air coming through a regenerative desiccant system that's got all the moisture sucked out of it at 75 to 90 PSI, and then you expand it to atmospheric pressure so it gets 7 times drier from there. So yeah, works like 1,000,000 bucks. I love it. Sips not all that much compressed air. If you don't have compressed air in your shop, you can buy a dental compressor. It's dead silent. I'm not going to give you any trouble. Awesome. Nick, thank you so much. That's Nick Parker, Director of Product. Now if you have questions or you want more information or a one-on-one presentation on the high low, then we're here to help. Shoot us an e-mail or give us a call. We're here to answer your questions and consult you and find out if this is the right machine for your business. And definitely act quick because they're going fast in manufacturing. Applications are filling up faster than any other machine I've seen, so definitely do it quick. Thanks so much for watching Compose the rest of your day. See you on the next video.