Our sales engineer, Paul DeWys, gave this presentation on Part Design for 3D Printing at a joint meeting of the West Michigan Catia / NX Users Group. Through the use of example parts he covers how to optimize your 3D CAD design for both prototyping and mass manufacturing on 3D Printing Equipment.
Transcript of this videos audio:
Okay I’m gonna be going off the top of my head. I might have to flip over my notes here and there. But we’ll see what happens. So my name is Paul DeWys. I owned DeWys Engineering and Fore Runner 3D Printing and this presentation is that today is designed for 3D printing. All killer, no filler. What does that mean? We’re gonna try and get right to the good stuff and get you into some information that’s actually usable for designers and engineers. Now a little filler. So I started DeWys Engineering out of my dorm room up at Ferris in 2009. My first CAD package was CATIA V5 this user group was a big part of me actually getting started. And from 2009 to 2016, myself and my team built up DE, we do a lot of things and obviously automotive, specifically shop floor automation and shop floor manufacturing gates. Everything from hand tools to robotic assembly equipment, kind of everything in between. Do some aerospace work, a lot of foundry work, a lot of heavy agricultural, mining construction. Then we also get into point of sale, point of purchase and picture development. Like you go into Speedway Gas Station, you see all the tobacco vending behind there, that’s a lot of stuff that we worked on for one of our customers.
New product development, you know everything from working on literally right now we’ve got a customer that’s wants us to look at making a recycling machine for cans to tailored down bathtub kayaks for American Girl dolls and everything in between. We’re very, very diversified kind of portfolio of work. And during that time, I used Select Manufacturing in Muskogean in my 3D printing vendor. [Ross Gates inaudible 00:01:52] was the gentleman who owned it, me and him became friends. He didn’t really have a succession plan so one day in May of 2016, I get a phone call from Ross and he’s like, “I’m done, I’m selling it, do you want it?” And I was like, “What? Your 3D printing business?” He’s like, “Yep.” And I’m like, “I guess, I don’t know, let’s talk.” So got together, fair deal, fair terms.
I bought the business from him, really bought the assets from him essentially. And we formed Fore Runner 3D Printing. So we started out with 2 SLA’s, an SLA 500 and an SLA 5000. [inaudible 00:02:28] watershed materials. And then we diversified into poly jet 3D printing and then literally two weeks ago, we installed an HP MJM machine and I’ll get into machine specifics here in minute.
So that’s a little bit about us. Okay so the big thing is like okay, design for 3D print, well I thought 3D printing was unlimited wild west, whatever I want, I can get. Well it is in most regards but you know it’s like anything, if you are designing the parts specifically for injection molding or specifically for stamping or specifically for production machining, you can gain certain economies and lead time and in part cost and in tooling costs, but with 3D printing you can gain those same things.
If you optimize your design for 3D printing, you can get isotopically stronger parts, instead of just traditional 3D printing where it’s strong, the X and the Y, but the Z orientation, there’s the layers so you might have cracks form in that orientation or heavy load. You can get around that if you design properly. You can cut a ton of post processing time. So you know if you are taking [inaudible 00:03:46] apart and we’ll get into this a little bit more, if you can mitigate the amount of supports that are required for it. You just cut all that sanding time and honestly, next to material costs, human labor is one of the biggest cost drivers for 3D printing parts. So anytime you can eliminate time on the bench, just is putting money back into your pocket.
So there’s some very simple things you can do to save yourself quite a bit on that. You can get added functionality and you need features. So 3D printing can do wild stuff that no other manufacturing process can really do economically. So you can design in features that can save you money and get you added functionality because this is a different kind of process. The biggest thing is you can lower your cost. You know a big part of what we do at Fore Runner every day is working with customers to say, “How can I take this from a 2,000 dollar part down to a 1,000 part?” And sometimes it’s as simple, “Hey can you split the part here and we can put it together afterwards? Or can we build it in a different orientation?” There’s always simple things that can be done that really don’t affect the end part at all but can save you considerable amount of money.
So the following examples are parts that we either designed specifically for the 3D printing process or parts that we helped customers re-design for the 3D printing process. One kind of fun one I threw on here, this is actually a steering wheel for a mafia themed classic car and so the guy found two models, or found a model of the Thompson sub-machine gun and had us essentially design him a steering wheel and then combine all the bodies and print it and then he finished it in a body shop and it looks pretty cool, so, that’s kind of a weirder one. But we’ll get into some more practical stuff.
So I’m gonna quickly though through each technology just so I can make sure everyone is on the same page. I’m sure some of you know how this stuff works, some of you probably don’t. I’m just gonna hit it all, just so we’re all on the same page. So we’re start with SLA, SLA is a liquid process, it’s like a two part epoxy, your A part is liquid and the B part is actually UV laser. So when the laser comes down and it hits the top layer of the resin, it flash hardens it, using that UV light and essentially it just hardens layer after layer and that platform just keeps dropping. And every time that platform drops, then you build your next layer on top of the layer underneath or the support underneath it, so that kind of bird nest material is actually support. So obviously you can just build in thin air or in the middle of the vat. You have to build support up to hold your part. So that’s kind of a quick run through of SLA.
So first example is this guy right here. And I’ll start passing these around. So this is a Blow/Vac nest, as we’ve kind of come to call it. Essentially the design challenge was to design a block that could route both high pressure air for blow off and strong vacuum to remove dust out of one part. We run a continuous feed of material across this. It has a powder coat going onto it but they want to make sure that the powder isn’t on one side of the part, where they put it on the other side. So this block essentially routes high pressure air to blow off the bottom and then vacuum away all that dust and leaves the top of the part untouched, so then it can go through the oven and actually cook the powder coat on. So we did some CFD analysis to kind of guide us on where to put the holes and how to orient them for the HPA blow off and also for the vacuum.
And it just turned into this real nightmare of like, okay we’ve got to find an EDM hole driller and it’s gotta be able to do fourth access moves and eventually what we kind of came to is, you know what if we print this in watershed and then we just put it into a stainless steel block essentially and the block will take any wear, but the watershed will allow us to route air and vacuum wherever we wanted it.
And we tried it and it actually works like ridiculously well. So this … there’s multiple of these in use on a shop floor here in Western Michigan and it’s working quite successful for the customer. So yeah, get into the kind of talking about specifics here a little bit, so the purple that you see traced through the block, is high pressure air, we’re just running shop air about 90 PSI. Kind of the sea green, [inaudible 00:08:23] vacuum, so essentially if you look in this cutaway right here, you can see there’s all these holes irradiating across the bottom of that and essentially they start from straight up and down the middle, and then kind of spread out in an arc. You can imagine trying to put all those holes in there with an EDM drill is gonna be pretty complicated and actually it turned out to be non-manufacturable when we really got into it.
But 3D printing really didn’t have a problem with it. And then the … you can see there’s these chambers right here, there’s four of those chambers kind of spread through the block, those are vacuum and with those, we wanted to come to a very specific mouth size opening to get the right amount of draw, to get that dust to break loose and then suck in and now just be aerated into the atmosphere. So with 3D printing, you don’t have to worry about … you know if we tried to do this out of a traditional C&C application, you’d probably end up having to bolt this block together from a bunch of different blocks and get seals and everything like that ’cause there’s no way cross drilling through a block would be able to create all the channels and everything like that, that we needed to move air and vacuum around so, the one thing to keep in mind with 3D printing, if you’re gonna have trap volumes like this is support. So remember how I said with SLA, there’s that bird nest type material that supports everything. Well you’ve got to eliminate that because it’s inside a part, you can’t get it out and it’s gonna impede your air flow or your vacuum flow.
So the rule of thumb is a 45 angle is self-supporting and most 3D printing applications, in this case, you can see the roof of the vacuum chambers right there, actually set those closer to like 60 just to make sure we really had plenty of angle and it should be very self supporting. So there was no build, there was no support in any of those build chambers. And then an arch, so one of the most basic Roman era engineering principles right? An arch is self supporting when [inaudible 00:10:22] held together. Well same thing applies to 3D printing, there’s no support needed for this arch. And same thing with this round hole down here, there’s no support needed in there. So everything is either a round, or a perfect circle, is an arch or has a higher than 45 degree angle on it so that we’re completely support free on the inside. And so yeah, it worked really well for our customer and this is just kind of a quick example of one of the things you can do. If you design for the 3D print process, you can do some really wild stuff like this.
And I’ve also seen people do stuff like this with fluid too. People do a lot of stuff with moving coolant around or anything non abrasive. If you get any abrasives in there, it’s just gonna chew this thing a part. But if you have a non-abrasive fluid and also not high heat either, SLA is really good at high heat, so you want to try and run room temperature if you can, that’s why air and vacuum are really great for this application.
So if you’ve got questions as I’m going, just interrupt me, raise your hand, anything like that and we’ll just kind of keep moving through these quickly. So that’s our first application. So that’s SLA. SLA is predominantly you know prototype based, that’s what we do mostly with our SLA machines. Shop floor applications are kind of few and far between just because of it doesn’t like heat, it’s not the strongest in the world, it’s a little bit, definitely more of the expensive side, so SLA doesn’t have as many shop floor uses but that is one of them.
MJF, this is brand new, this is from HP. This machine has been out commercially about two years now. So like I said, we’ve been running these parts for about a year and we just got our machine here a little bit earlier. So this is a nylon process. These parts here, this part here, essentially the machine lays down a 3,000 layer of nylon, then it comes across and it sprays the black fusing agent and then it sprays a clear detailing agent in the same pass around the outside and it’s a high temp, internal environment so the powder is probably within 25 degrees of it’s fusing temperature. So when that black fusing agent goes on there and then a big powerful light goes over the top of it, it fuses that layer of nylon but only where you apply the fusing agent. The detailing agents there just to give you nice crisp edges and corners.
I’m going to pass around a whole handful of different parts here, you guys play with. And kind of go from there. And I’ve got tons of other stuff up here too if these don’t make it to you. You can come up and see me after the presentation. There’s a handful of stuff. All right, so MJF, so this is a citric nylon process. These parts are 99.5 percent dense, air tight, water tight, it’s pretty much equivalent in performance to machine [inaudible 00:13:34] nylon. It’s very isotropically strong, X, Y and Z because of the fusing process that they use. So we see a lot of applications for this in our engineering business, a lot of machine build, [inaudible 00:13:48] arm grippers and hand tools, nesting details. We’ve been working on all those for some customers and having some good success. So this first one is a safety cover for a high speed ratchet, so the customer was getting these machined out of Delrin. And they just kind of ran into the situation of lead time on the CNC, and because it’s ratchets they change constantly. There’s not enough volume to justify a mold because they’re constantly changing what ratchet to put it on so they would have to have a warehouse full of molds to cover every possible ratchet size and geometric that they might need.
They’ve been machining them but lead time are getting longer and longer. The spindle … in this case the lead time is getting more and more expensive. And so they came to us and said, “Hey, is this something that we can maybe 3D print?” And with this MJF process I mean we can fit 400 of these in one build chamber. So we agreed to it, but the problem we ran in to, this is the original design that you see right here. It’s a complete circle. There’s no splits or anything like that in it. So when we went to assemble them, the nylon would actually just shatter because it’s not quite as elastic as the Delrin is. So we kind of played around with it for a little while and we realized, hey if we put some petals, some cuts into these parts and kind of petal them, that allows the MJF part to just flex and then snap right back. And still has the same performance characteristics.
Holds on there really nice but the parts don’t just shatter when you go to assemble them. And that’s one of the things to keep in mind when designing for an MJF or a SLS type process. You know your materials aren’t gonna be made quite as elastic as a Delrin or a UHMW. So the trick is, petal them where you can and give yourself essentially turn these into four individual snaps instead of one solid ring. And we found that that worked really, really well for this and some other applications.
Another thing that’s really cool about you know 3D printing is labeling. So you can label injection mold, you can stamp labels in, you can engrave labels in a CNC machine but here it doesn’t drive cost. I mean it’s literally, if you say, this is the part or [inaudible 00:16:15] the part, no problem. It’s three clicks and magic and I’ve got every whatever label I need in there. So the customer really liked that too, we started printing the size into every single part and they really liked that for when their guys are going to assemble it. Because inevitably, a guy grabs a handful, a handful, a handful, he dumps them on the table and then is like, “Oh shoot, which one is which?” So just printing the label in there was a big deal for helping them on the shop floor. So yeah, so that’s another example of taking a traditional machine part, transitioning it over to 3D printing.
End of arm gripper fingers, so there’s one of them going around right now and I’ve got a couple others up here. These are really, really pretty cool. The issue that our customer was running into is, they had really, really tight space that they had to load parts into it, an injection mold for overload, so inserts. And due to the number of grippers required, the end of arm tool also needed … had a weight issue there, their robot didn’t have the capacity to do a whole bunch of aluminum gripper fingers. So they were looking at machining them out of nylon and you know that was gonna work but the only other thing was, it was such a tight fit and they had to have suction cups in the ends right here, there wasn’t a lot of room to actually channel the airline around the outside of the gripper finger.
So we kind of said, “Heck I wonder if we can print the air line inside the gripper finger?” And so the one going around, you’ll see there’s a suction cup inside and there’s a fitting on the outside because it’s actually 3D printed air channel through the inside of the gripper finger, so now when it reaches into that really tight space, you don’t have to worry about an air line getting caught on anything, the air line is actually coming out, outside of where it’s reaching into the mold.
So this is kind of a cut away of what I was just talking about. In this … the fingers that are going around, you know like I said, it’s center nylon, it’s actually and this is where my notes would be helpful, but off the top of my head, I think the aluminum version of this, we’d turn this aluminum into solid works and it was .45 pounds, I’m gonna look real quick, .41 pounds and the MJF version was .18 pounds. So there’s a pretty significant weight savings with going with nylon. Nylon is good up to I believe this is good for up to 300 degrees Fahrenheit so we didn’t have an issue of you know these getting all floppy on us and the mold and it’s getting ambient heat put into it every time it reaches in there.
And that printed air channel worked out really, really slick for moving the air through there and then if you look at this as you go around, I mean if you wanted to do this, you know this could be simplified a little bit but if you want to do this on a C&C machine, you’d be like probably I think three set ups to do this. And then you’d have to be gun drilling into it to put your internal air line through it and then you’ve got to put the plugs in, so it can be done, definitely can be done traditionally on a CNC machine but I can print this out the door with the fittings into it and sell it for a hundred bucks a piece. So the part that’s going around with fittings in it, that’s a hundred bucks out the door.
So you know so this material, you can drill and tap it. So in this case, we ran pipe taps into it, put our fittings in there and from EMI and it holds vacuum, works really, really nice. We have … I haven’t done it yet but one of our vendor who installed this machine for us, they’re over in Detroit, they’ve been doing a lot of experimenting, they’ve run these up to 125 PSI and haven’t had any issues, you know so it’s pretty stout stuff.
Obviously if you make it small enough, eventually, it will burst, but if you make a beefy part, it will hold up to some pretty good pressures. So you know end of arm gripper fingers. Oh and the last thing, there’s a texture that we put in here to grab onto the parts. In this case, you know the only way you can do that is traditional manufacturing would be like an [inaudible 00:20:18] sinker type operation. And again, you can print your textures right at. I was just at IMTS yesterday and went through the HP booth and they have customers from all over the world that they grab sample parts and bring them in. And some of the printed in texturing that people are doing for decorative, is just wild. You know if you’ve got some time, check out the HP website. I know they have some sample parts up there with these printed in textures that is just like crazy looking. So that’s our gripper finger example.
Another interesting one and this is one where we’re still like literally in development on, one of my buddies is a photographer and he also has a Amazon operation that he runs. Sells a lot of custom made camera gear on there. And he came to me when I got this machine and said, “Hey, you know I would love to do customized lens caps, lens caps for people. So essentially I can put your name, I can put your log, whatever you want into your lens cap.” And he said, “But you know I gotta keep in 35 bucks on my Amazon page to be able to really make these things sell. Is that something that you can do?” “I don’t know, let’s try it.” So we did. And this is one of our 1.0 versions, I’m not gonna pass it around ’cause it’s all broken and it’s not really that great, but I’m gonna show you what we did do to fix it.
So this is really neat because when we’re talking about trying to hit that 35 dollar price point, I really need to be able to sell it to him for 10. And you know if he’s gonna order them a hundred at a time, so we’ve got the quantity that we need, you know and to customize it with different names and stuff like that, again it’s 10 seconds a piece and magics and we have an intern whose life I’ll ruin with that job, so it’s not like a big deal for me.
So anyways, so the trick was, well the one he gave me to model, it had, it was a three piece design and it had a two springs that they put in there to get that spring action so it would lock into the end of the lens. And I’m sitting there going, man, if I gotta buy springs and I gotta together a little assembly station and I gotta pay someone to sit there and assemble these things, it’s gonna start to add up. So we’ve seen other people do it, so we thought, hey, let’s try 3D printing that says a complete assembly with the springs built in as printed features.
So this nylon HP does have a glass filled version of PA12 which is this material, this nylon. We are not running glass filled version right now. We’re just running straight PA12, so it is, I mean these … you can try the one you got out there. You’re not bending this, this is very stout, but if you keep the form factor thing enough, it is pretty springy.
And so what we did, is we experimented and you can see this gap right around here which is all the way around the part. We found .2 millimeters, if you leave .2 millimeter gap to .3 millimeter gap around your assembled component, they’ll print, they won’t fuse together and then you can break the powder out just be actually moving them a little bit. One of the reasons this is broken is you’ll notice this area is solid and up there, that area is open. We didn’t think about, “Hey the powder has to get out of this after we print it.” So we printed it and I went like that and it just went crack, and I went oh well that didn’t really work.
All right, send us another one. And that’s the other cool thing about you know proto production is what some people are calling it. You know you print three, four different designs with your springs because you can take a spring design you traditionally use in a traject ion molding, that’s a really good starting spot for 3D printing.
But then you can kind of tweak it from there and it’s all about iterating these springs right? Well you know the nice thing is, you throw four of these in the machine, you print them over night, you come in, okay that one didn’t work, that one didn’t work, that one didn’t work, oh this one worked pretty good. Iterate on that two or three more times and then literally next week, we’re gonna do our first production run of a hundred of these and we’re able to sell them out the door to the customer for 10 bucks a piece and he can go and retail them for 35 on Amazon.
So you can print assemblies, you can print springs into your assemblies, like I said, beaten a dead horse, I was at IMTS yesterday and some of the stuff I saw, people who were at the EOS booth which is an SLS machine, kind of similar to this. HP booth, there’s guys printing full gear drives and assemblies, totally enclosed, they put them into a parts washer, wash all the powder out and go. I mean they’re doing gear boxes out of this stuff with very minimal openings, just enough to get the powder out and then they plug them. So you know really, I hate to sound all full of hyperbaly, but like if you can think of it, there’s a good chance you can actually get it to work with MJF or SLS technology. So kind of a neat application.
So now we’re moving in FDM, this is like kind of the mainstay, this is the maker bot, this is the [inaudible 00:25:37]. You’ve got your spools of material, you feed them through, and you’ve got a support material and a build material. And you know these machines, are really, really good in our business, we have one and we always kind of use it for a lot of hand tools and a lot of more of kind of the quick and dirty models. To kind of get up and running, so here are some hand tools, I’ll get into in a minute. Best thing about this, it’s cheap. I can go on Amazon and buy a role of material for 27 dollars and so my customers, they need hand tools, they need a lot of them, they don’t want to spend 1,000 dollars a tool.
They want to spend a hundred bucks, 50 bucks a tool. And that’s what we really like this for, is it gives us the ability to really iterate really easily on hand tools and then to make them at scale for a reasonable cost. So first application here, is a tape hand applicator. So the challenge, design and 3D print a series of ABS plastic hand tools that can be used to guide the application of tape onto windows during assembly. Our customer supplies all the tape to Anderson Windows, down in Mexico and they just … these factories are just full of people and they tape all the windows together.
There’s not a screw or a bolt in any of those windows. It’s all double sided sticky tape. You go look at a skyscraper, that’s the really crazy part, ’cause I’m like, “Oh Anderson Windows.” And then he’s like, “Well next time you go downtown Grand Rapids, look at the Marriott or look at the Amway Grand. All that glass is held in with double sided sticky tape, that’s it.” Now this double sided sticky tape, I also know for a fact that if you make the mistake of setting a roll of it on its side on your desk and going home for the weekend and coming in. You’re gonna peel all the laminate off your desk to get that roll off. So it’s very, very high end tape. But still it’s just tape, kind of crazy.
So the tools were gonna be used on daily basis, so they needed to be robust. They needed to be able to stand up to wear and tear, and then there’s a large range of window size and designs that coupled with the cost to set up CNC machines for each size tool required and the need for the finish cost of the tool to be less than a hundred dollars each, made traditional machinery too expensive. They had been machining these out of Delrin, and it’s actually this style is what they were machining out of Delrin which didn’t have the rollers in it or anything like that. And they couldn’t, because of the variety of sizes that they needed and the fact that they only needed 10 to 12 of each size, they could never get the cost really low enough to make it super competitive. Because at the end of the day, the tape company gives these applicators away to help sell more tape, so they want to make their customers happy as they can but they don’t want to pay anything for it. So it’s kind of one of those catch 22 situations.
So we got involved and we kind of said, okay let’s look at this and see if FDM printing would be a way for us to do this where it’s cheap enough to get the price point that you want, but flexible enough to cover wide range of sizes and it’s low enough in volume where you don’t have to buy 200 of them to hit the prices that you want. So what we came up with and then once it … it’s kind of like most things, once we sat down with a customer, talked with them about the flexibility that 3D printing provided, he started thinking, “Well what if we put rollers in because after they take and they take the tape down, they gotta take a little roller and sit there and go across it and wet the tape out.”
He’s like, “Can we combine the roller with the tape applicator and save them a whole set of movements?” And so we’re like, “I don’t know, let’s try it.” And that’s what we did, we tried it and it worked like gangbusters, like they sent one down to Mexico and a week later they got an email back, and they’re like everything needs to be like this, don’t ever send us one of the old ones again. Which was great, our customer is happy, his customer is happy, so then the next challenge kind of became okay, let’s figure out how to kind of make it robust, a design that’s robust enough.
So one of the things you’ll notice on these blue ones, is the sides are flat and they almost … they have a different finish than what you see on the rest of it. The rest of it has that kind of traditional layered feel to it, those sides are flat. And that’s because when we designed this, we said, “We want to design this with as little support material as possible.” Because remember support material just drives time and time drives cost. And we’re trying to keep the cost down. So we designed these to be printed as you see in this picture right here, laying flat on the bed, print everything up and literally only support we have to break out is in these counter boards here and in these counter boards over here, we have to dig the support out of there. But that’s not too bad, two minutes with a screwdriver, that’s all cleaned out, go in there with a dremel, sand it in a quick second, you’re done. So you know three minutes per part for [inaudible 00:30:42] benching them, that’s inconsequential because the cost of the part.
You know another thing we did with this, is when we print them straight up and down and this is kind of across the board for all 3D printing, if you need a truly round part, or a round part or a round hole, and or you have dowel holes and you need them round and locationally accurate to each other, you always want to orient your part versicle in the ZX, you want that hole facing up, that gives you the best chance of not having to dig support out of the hole and also if you build that hole on its side, not really round, it’s more kind of egg shaped when it’s all said and done because of how the layering works and how the part kind of settles as it prints.
So your best bet is to flip it straight up and so that’s another thing we did, is we had all of our holes pointing straight up into the Z so they were nice, true, round holes and we didn’t have to worry about them being egg shaped because you know we have this roller, we’ve got that kind of bluish roller in there, you know we want that to run nice and smooth and we found when we stood them up, the inside would be rough and it would just kind of hiccup and get really messy when it rotated, it didn’t do a really good job. We put them on the side, rotates really nice.
We had bearings in there originally, we wound up pulling those out, with just running on the 3D printing part, it was smooth enough. Another thing to think about in orientation, is with FDM, you know if you have an angled surface like this up in the Z, and you’re printing up like this, you get a stair step effect in that. So you can see we have these ramps on here, we wanted those ramps to be super smooth all the way up where the tape is [inaudible 00:32:31] through. So by laying it on its side, now all of a sudden, your printer is tracing that ramp. So there’s no stair stepping, it’s a true, smooth angled ramp the whole way. And so that was another thing that we found that made a huge difference in the kind of quality of these was doing that. So kind of key take aways, make your holes facing up wherever possible. Kind of think about how your layers are gonna be oriented and how your parts gonna be used, you kind of optimize things that way.
And then also last point on these, they need to be strong, so we printed these at 85 percent dense, so there’s a lot of density to these parts, you know typically if you go and get parts FDM printed, they’ll run them around you know 13 percent to maybe 25 percent infill and those are … they feel real good, they’ll work good as models, but you start using this hand tools, they can start to break down on you, so you want to get that density up.
This is a weird one, dog head MRI fixture. You know in my career I’ve had some really weird phone calls. People asking me to do some really weird things, this was by far one of the weirdest. We got a phone call from a lab in Muskogean and they were like, “Hey you’re in that 3D printing business. We need a fixture.” And we’re like, “What do you need a fixture of?” And they’re like, “Dogs, an MRI machine, so when we sedate them, we need to lock their heads into position so that they don’t …” I guess when dogs are sedated, kind of like your dog at home, the leg just starts kicking for no apparent reason when it’s knocked out. Well the same thing happens when you sedate them, so the dogs in the MRI machine kind of moving all over the place and that doesn’t work well for an MRI, so they want to sedate them and then fixture them into position so that they can take good, clean MRI images and that’s that.
This is not our lab, this is just a goofy picture I found online to prove that yes, dogs, you know what’s really crazy, when I typed in animal MRI, racehorses, holy smokes, there are custom built MRI machines just for X-raying the knees and ankles of race horses. So on your lunch break today, Google racehorse MRI and be amazed.
So yeah so the fixture needed to be very flexible too because dogs come in many shapes and sizes, so it’s not a one size fits all. So it looks a little medieval but it promised no dogs were harmed during this fixtures. They’re trying to help the dogs. So they used these two to kind of position off of the dogs ears and then these two flip into it’s jaws and one of the things that was interesting, those targets are actually have, they’re loaded with mineral oil because mineral oil glows super bright in an MRI and they use that to align various shots off the MRI machine and then some software they’re using, they had to have these glowing targets.
And again, they were very adamant on the positional accuracy of these cylinders in these four different areas need to be as accurate as possible. So we actually built these standing straight up on the bed of the FDM machine, they were about that tall but you know you’ve got a perfectly round cylinder in the center of that part, we just pour mineral oil in there and then cap the end of it and you’ve got your mineral oil target.
So again, orientation was really important on those two. Build strength was really important too so like the frame here around the outside the kind of horse shoes on the bottom and then this back piece and those blocks, we actually printed those at 100 percent density, so they’re as dense as we can get an FDM part and that just makes them insanely strong. So again, when you’re talking FDM, be thinking about your strength requirements and let your … either your internal guy or the service bureau you’re working with, let them know, hey I really need it strong in this orientation, I really need it strong for these types of loads. And there’s things we can do to improve the strength of your parts.
But also the flip side is true, if it’s not a real load bearing part, let us know that too because we can drop density and that will save you time and money, if we can make the parts with a 13 percent infill, that’s gonna cut our build time down to … on something like this, you can literally shave 40-50 hours off a build time just by driving your density from 99 percent down to 25 percent. Another thing is, so it’s an MRI, can’t have any metal in an MRI, that’s why we originally got the call, but you know one of the things, you see these green fasteners all over the place. Well we still had to [inaudible 00:37:14] to adjust to things, so we needed stuff to be threaded together but we couldn’t … typically when we 3D parts, we’ll [inaudible 00:37:22] holes like if you’re gonna be bolting something into a machine, you might be taking those in and out, in and out, in and out. We’ll heal the coil or we’ll epoxy insert with brass threaded inserts into the part to make sure those threads don’t just burn out.
Well we can’t heal a coil this so what we came up with is, we’re gonna … first off, if you’re going to use 3D an actual tap threads, try not to go below a quarter twenty. Obviously I violate that rule on this one here but this is a one time assembly, like you probably aren’t going to be pulling this thing apart and putting it back together a whole bunch of times, but if you are gonna print 3D threads and you are gonna be taking apart and putting them back together, don’t go below a quarter twenty and half is even better. If you can get it to the half inch size range, those are really, really good when you tap them in 3D printing.
You can 3D print threads too but that gets really, really tricky and it’s a big cost driver, so we always like to say, if we can just you know drill and tap afterwards, we prefer to do that. You know we’ll … you know if we’re using I probably shouldn’t use this as an example, if you’re using solid works and you’re in the hole [inaudible 00:38:36] you know you can literally tell, hey make this hole, the tap drill size and then all you gotta do is literally take a tap, spin it in there and you’re done. But in this case, we kind of had an epiphany and said, hey if we’re gonna tap into the FDM plastic, let’s actually buy plastic screws to go with it. ‘Cause obviously you can’t have metal but plastic thread on plastic thread is gonna last a lot longer than metal thread on plastic thread. So even you guys will have applications, if you are going to be tapping these parts and putting them together, if you can get away with using nylon screws off [inaudible 00:39:14], you know try and go that route ’cause your thread is gonna last a lot longer that way. So kind of … the takeaway from there, heal a coil or insert if you can, if you can’t, drill and tap, don’t print the threads and then try and use plastic fasteners.