Description
Key Learnings
- Discover the significant differentiators between additive manufacturing (3D printing) and CNC machining
- Discover decision factors that go into choosing your end-manufacturing process
- Discover the benefits and “sweet spots” of additive and subtractive manufacturing
- Discover design-specific cost drivers between machining and 3D processes
Speaker
- Greg PaulsenAdvanced manufacturing expert, 3D printing advocate, and application engineer at Xometry.
GREG PAULSEN: Hey, everyone. Thank you so much for joining today's webinar on Should I Print or Machine This Part, learn which process to choose and why. Today, we're going to be talking between 3D printing and CNC machining. Our goal is to discover some of the significant differentiators, the strengths and trade-offs of these different processes.
We're going to talk about decision factors, things that I may be thinking about while I'm designing for a part and when I'm in a product development phase of making that part that will help me choose which manufacturing process to go with and look for sweet spots, look for things that really identify, this is a great application for additive manufacturing, or this should definitely be machined. And I'll work through that. And I also have just a great summary just talking about how cost is driven different ways depending on what process you're going to choose.
So who am I? My name is Greg Paulsen. I am the director of application engineering at Xometry. And I work in the world of help me choose, the world of manufacturing process. In fact, I've actually worked on over a dozen different manufacturing processes ranging from machining to molding to multitudes of different additive processes being someone who's worked directly in applied additive manufacturing for about the past 14 years or so. And I'm always happy to geek out and talk shop as well. So this is right up my alley because these are conversations that I have every single day.
And where's my stomping ground? Well, my stomping ground is Xometry. We are a global on-demand manufacturing marketplace. So we are one-stop source for manufacturing parts in all these different processes, many of which I will be speaking about today, but that would be additive manufacturing so a multitude of different process from powder bed, SLS, HP Multi Jet Fusion, for example, filament base, FDM processes, photo resin bases like Carbon DLS, SLA, and so much more.
When I talk about subtractive, it's going to be the gamut from micromachining to Gantry milling to turning to EDM. You name it, it's something that we've had experience in reading drawings, understanding the specifications, and how to qualify what you make at the very end. And the other things that we offer at Xometry is tooled stuff so molding, casting, die casting, insert molding, overmolding. And again, these are all different projects, and we know the strengths and trade offs of each one of those as well.
And what makes it great for you as a user, as a designer is that there's a free site where you could go and actually kind of do some sanity checking. You could go to Xometry's site, configure your part, get pricing and lead times right away. On the right-hand side there, you can see our modify part area, and this is a highly spec'd out CNC machined component, actually a pretty big one, a 30-inch one. But you can really get down into the details, see how your pricing, lead times, update.
And from there, it's actionable. You could just press Buy. So if you're like, that's great, let's go for it, boom. We could get that part into process. Because we're connected with over 5,000 manufacturers, it allows us to do work consistently and in parallel. So you could give us 100 jobs today. We could get them started tomorrow. And it's really fantastic regardless of the process that you choose for your manufacturing step.
So let's get down to business, though. When to choose CNC or 3D printing? And again, the goal here is to really help you identify key factors and even learn and understand a little bit more about these different processes. But I always like to start with mental models starting with my paradigm shift. What do I know? And where do I need to kind of break what I know and try to understand a new process here?
So let's just start talking about traditional manufacturing. This is starting with something larger, a billet of material, a large, flat sheet, and reducing it in some way to the shape of the final part. So that could be cutting, milling, drilling, bending, et cetera. And I'm forming that larger thing, the thing that's spec material, into these smaller shapes. And you probably just talk about a CNC machining or CNC turning or sheet metal fabrication.
This is what we know. This is what you see around you. This is something that you typically have been taught about in school. So it really is where your paradigm starts. You're looking for machined remnants. And even when I talk about what you know in other things around you, obviously, there's injection molding. That's still from a machine tool cavity probably.
Stamping, it's a machine die. All the stuff starts with that subtractive what I'll call traditional manufacturing. And I'm going to say there's nothing wrong with it, and believe me, we do this all the time. But I will say that additive is going to be a little bit different there, but we'll get into that in a moment.
So CNC machining, what are these typical results when I order a CNC machine part? First off, I'm cutting these basically with a high velocity spinning tool that is going against my built material using some sort of computer aided manufacturing program. So you can see the spinning end mill here that's going against and creating a part of this form. In this case, it looks like a tooling form here, but it's relatively smooth. I'm removing material. I'm chipping it away very quickly, and I'm leaving typically a 125 Ra-- that's in micro inches-- surface finish as machined.
I am using a spinning tool. So spinning tools all of a sudden don't have square edges. They're always circular so it means if I'm doing a internal cavity, I need to add some radii on the inside to make sure that that feature can be made. So if I have a right angle, for example, on the inside, it requires a different type of process, such as broaching or EDM to make, which could be a giant cost driver there. So best practice is always assume internal corners need to be rounded.
I'm also using this tool. The tool's not going to go all the way through and then just curve at the very end just to go into that one little spot. You're working with line of sight directions. When you're designing for this, it is your CAD commands. Again, we're talking what we know. Extrude, cut, what length, what distance? Your CAD is designed around what we know, these traditional paradigms. So it's pretty familiar information.
And what drives the pricing of this is typically the setups. So if you want to have a troubleshooting friend, know machinists. Every new job that a machinist works with is a new problem to solve, how to work hold it, what operations do I need to set up, but that's their job, their labor. That's their set-up operations. That's work ahead.
Regardless if I'm making 1 or 100 of that part, I need to build the setups for those operations first. So on low volumes, those setups are encompassed in that part itself. As you build multiple volumes, you're going to find that that setup fee amortizes through, and that's pricing behavior of machine parts.
Machine overhead, these machines are not cheap typically. So the time on the machine, how long does a milling operation take per operation, that adds up. And that'll be part of the pricing. And then secondary processing, am I installing inserts? Am I putting finish? Is it getting anodized? Is it getting plated? What type of QA? Is it just manual tools or CMM inspection? All that's going to be part of the cost driving for this work.
But it comes with fantastic benefits. Because it's what we know, because of how characterized CNC machining is, you are already really set up well for it to manage your expectations. This general tolerance for CNC machine parts distance-wise is typically plus or minus 0.005 an inch-- so that's about the thickness of a sheet of paper-- and a delta of distance on any feature set. The feature there, I already have those smooth surfaces as we talked about.
And the characterization of the materials, how well they're classified, alloy, temper class, you don't question it. When you go and order a 6061 T6 from any vendor, you're not thinking to yourself, but is this 6061 T6 different than this one? You have a high confidence level in what those alloys do, what they produce, especially if you're simulating and trying to figure out some of these engineering challenges.
You really have a database to work with already. And of course, it's what we know. It's traditional. It's the foundation of most manufactured components today so extremely established. And it can be quick, but sometimes it does take longer, depending on the operations and setups there.
So for example, Phase Four Aerospace, a Xometry customer, they use CNC machining. They like us for our CNC machining. And here's the reason why. They're building microsatellite propulsion systems, and they need something that is highly precise. They don't want to wonder what the material is going to do. They need to actually classify and characterize it.
They need specialized finishes so a type of chemical conversion coating on the end. There's even you can see some silk screening on that logo piece there. And they just want the stuff to be put together. They have a technical drawing. You QA against a technical drawing. You get your parts.
So our CNC services, we're talking about a part that is going to be a space flight part, something where you can't, hey, take it back down and regrind this area or do this. You can't maintain a part when it's up in space. You want that level of confidence. They're choosing CNC machining for this.
So it is a really, really powerful production method because the high amount of customization you can do and because of that characterization. We know what that material is to start with, which is really something that when you're looking at very high stakes information and you don't want a research project before it, it's a fantastic approach.
So on the contrary, let's talk about 3D printing. When we're cutting away parts, what happens when I grow them? So 3D printing is the process of joining materials to make objects from 3D model data. So it's not just an extruder pen. It's actually something that's being dictated by your CAD model to make those shapes.
It's a layer by layer process. So I'm typically fusing those layers together either by a heat or some sort of photochemical reaction to get those materials to connect to each other left and right so on the XY plane, as well as the layer underneath, that z plane, giving me that 3D of the 3D printing.
Most of the time, we're calling 3D printing just kind of when we're having conversations on an industrial end, we tend to lean towards additive manufacturing more as a classification of the manufacturing technologies. And something to note is that it's not just one thing. It is an umbrella of different technologies available.
In fact, when I talk about 3D printing, in 3D printing at Xometry, for example, I actually have eight processes. I have seven listed here-- Binder Depth Metal. I realize I need to update this slide here because it was a recent addition to our additive classes that we instantly quote online.
But each one of these has a unique, differentiated method of making a part in three dimensions from either a thermoplastic, a photo polymer, or metal. And each one of these has its own strengths and trade-offs. So I just really want to note that 3D printers are not just one thing. It's a whole family of different technologies out there.
What I live in is industrial 3D printing. So I'm going to be speaking mostly to industrial-level technologies. And in general, though, I am still growing these parts on a layer by layer basis or infusing together, which means that unlike something where if you take an Exacto knife and you cut away, or if I'm whittling a piece of wood, for example, you're able to make a smooth cut. And that's where you get that smoothness on a CNC part.
Now, if I'm putting something together and building up, my surface finish is different because I'm forming that part. I'm growing that part from a base outwards. So depending on the medium used to form, you're going to have different levels of surface roughness, if you will. So liquid photo polymer resins, for example, when I'm 3D printing will typically have the smoothest natural as built finish.
Powder bed fusion platforms like Selective Laser Sintering, which those machines on the left there are Selective Laser Sintering machines, will be typically kind of a matte, sugar cube like finish. And filament-based processes like FDM or even that desktop filament-based printer tends to have more coarse layer lines or what we call stair-stepping on layer by layer because think of it as a bunch of flat patterns kind of fused on top of each other.
The beautiful thing about this, though, is don't think cutting something down and making squares. Think about growing something out. Think about coral reefs, stuff that can actually have more of an organic nature to it. So complexity tends to be pretty cheap when you're able to take from a base unit and start to move feature sets outward on it. It's something that you can really uniquely take advantage of with 3D printing, for example, for lattice structures and other lightweighting features.
And the cost drivers, though, is the material. So more stuff on the part is work, which is ultimate of CNC machining by the way where the less stuff I have to do, the less work, it's cheaper. So it's kind of retaining material is cheaper in CNC machining. Machine overhead, again, these machines on the industrial scale could be hundreds of thousands of dollars. They do have higher overhead rates.
So that time on machine can be very significant in the price of the part. Sometimes more than that, actually often almost always on an industrial scale more than the material cost. And then secondary finishing, and I vapor smoothing this? Am I doing manual labor against this part to bring it to a different surface finish? That's going to be a very large driver on a part by part basis in industrial 3D.
But here's the benefits, right? It is the leader for prototyping. I can get parts in one to three business days. It's very geometry agnostic. You don't need to set up ops and tooling. You're able to essentially provide a CAD file, which is a bare minimum to get into 3D printing because it's interpreting that 3D model to do its darndest to make that shape for you. And you're good to go.
So when I'm prototyping parts or doing low-volume production, I can often save a significant amount with a 3D printed part because of the lack of tooling required. So I'm able to really start and utilize right away just off the stomping ground. This is also fantastic when you think about that complexity where I can all of a sudden do free-form geometry.
I can combine geometries together to make all-in-one assemblies. So I could do a lot more creative work with the additive manufacturing component versus CNC milling just because of the nature of this design. You know how we talked about CNC machining needs that line of sight feature? Not necessarily true with 3D printing.
So a case study here, I don't have Phase Four Aerospace. I have Dixon Valve. Dixon Valve, they use multiple manufacturing processes, and they needed fixtures. So this is something on the assembly and tooling line. And actually, we find that this happens, this is a challenge we see all the time. And additive manufacturing usually is the first to call for these types of applications.
So they needed some fixtures to put onto their platforms that had to be a very customized geometry to help on the post-machining of investment cast components they had in production. They actually were able to use our services because in this case, they actually needed metal.
So ignore this. This is actually 3D printed as well, but I think that's their own desktop machine. But this part right here is direct metal laser sintered aluminum, and they were able to put features, little protruding nubs, little overhangs, and other features that may be very difficult to machine, and they just needed to be purpose design.
They had this one job, the job that it was meant to do, and they were able to design around that job and make the perfect part for that function using a lot less material in the design as well as without the need for setups. It's a direct print experience. So they were able to utilize us for in their tooling jigs and fixers category. And again, it's a best friend when you talk about tooling design and engineering.
So we've talked a little bit about design, but let's go deeper into this because again, we know CNC is what we know so I'm going a little bit deeper onto that 3D printing realm.
So the first three paradigms or really the big three that you need to think about is bottom up. I am fusing my material usually on something underneath, whether that is the build plate, the substrate that I'm building my part on, or part of that part itself because if you stop a job halfway through, you just have half your part there. So I need to make sure that I'm fusing my layers on a layer by layer basis. There's some rules that come out of that, some efficiencies that you could gain when you think about how to grow something out following design guidance for different 3D printing processes.
The other thing to think about with this is that the machine is interpreting your design, but the outcome also depends on your design and your intentions and what you're looking for. So a 3D printer, if I have one edge that is a plus 0, minus 4, for example, on CNC machining, I'll be able to slow down as a machinist and go and measure that edge, do a little bit of offset, measure it again, and tune it down to get that machine feature.
3D printers work a little bit differently. They make near net shapes. Some are more technically accurate than others, but it's a combination of the process, the material used, and that CAD geometry, and a little bit of orientation as well on what the output is. Something that you may find is that guaranteed tolerances, they're usually pretty close, but they're not 100%, especially on the first print because a lot of that CAD geometry that you provide may work against you depending on how well it's designed for the process.
For example, if I've designed a brick for 3D printing, there may be thermal processes that may shrink it down more than the tolerances. So it's really important to understand some of the guidelines of these processes as well as understand that the machine is working on a net shape basis. So you get general tolerances, and a lot of the small detail changes for fits among your parts have to do with CAD offsets, CAD intent up front.
The other thing that CAD intent's very good for is understanding that as I'm growing these parts, especially in parts that are not powder bed where the parts are suspended in powder, I need support structures. Support structures are built as sacrificial structures that will take overhanging features as the part's being built and essentially give them a place to deposit.
So if I'm building the letter T, I could build from the base up the middle shaft of the T, but as soon as you have those two protrusions going out on each side, I need something where that plastic stands on. If not, the plastic will just sag down, and I'll get more of a floppy umbrella look. So building up supports from the bottom to the top alongside that shaft until I get to that layer allows me to place material down. And then after processing, I remove those supports manually or sometimes automated, and that gives you your final result.
Additive is an umbrella. So I had that word "choices" because help me choose is a big part of my job. This is not the main point of this webinar so I'll be very brief on this, but we'll also have the option to take a look at this, and you can pause and also look at the documentation here. But we do have some really good tips, but it's different than what you think about with your traditional manufacturing process when you're thinking machining.
It is about application. So clear, smooth, chemical resistant, these main characteristics are going to help better than understanding, do I want ABS, or do I want PC, or do I want ultem, or do I want nylon usually because you can make additive components behave very differently depending on the processes. Even things like nylon, nylon is available in machining, in several additive manufacturing technologies, and they're not all the same. The properties that come out can be very different.
So usually when we're looking at choosing different processes, we're talking about goals. Is this a cosmetic or aesthetic prototype? Maybe I need something clear like this part here. So I could use something like our SLA process with a quick clear coating, which is kind of a clear coat after the formation of the part.
If we're looking for something more function, then you have very simple down selects when it's like, do you need metal? That brings it down to two categories here. Non-metal brings you up to more. Rubber-like, OK, I got a few categories here. Non-rubbers brings you down to a little bit more.
But you'll find that depending on what's really important about this project, you'll find a smaller down select of materials available. So for example, if I'm looking for something that has a higher heat deflection, I may choose Ultem 9085 or Ultem 1010, which is a FDM category, or digital light synthesis has cyanate ester, or CE material, which also has a higher heat deflection to it, I think around 231 Celsius.
The other thing to note is that platforms, not all of them are the same. I've ran into the customers before where they absolutely loved the part that they made in say, for example, Carbon DLS, but then they're like, but this next part is 24 inches long, and some of these machine platforms for those processes just aren't up there yet.
So we may have to move to another process like fused deposition modeling or stereolithography to get those specific geometries. So on the bottom right here, you can see where different platforms typically cap out. There's a big asterisk there because there's always exceptions to the rule, but it's something good to keep in mind.
And we noted this before, but just from an analogy, when I talk about powder bed processes, when I talk about photo resin processes, when I talk about filament-based processes, a good rule of thumb is this pen analogy that I have. So powder bed processes usually are heat fused together.
It can get great detail, but if you start to have really small details together, you may get some thermal bleed, which is basically that laser, that fusion process creating heat around the parts that are actually supposed to be fusing, but they're also close enough together that the radiant heat is creating a little bit extra fusion there.
Not much, but it may change your tolerances by a couple of thou, and that could be a bigger deal. If you're trying to fit a square peg in a square hole, that could be a big issue for you. So I like to use that Sharpie fine point because, again, it's a great marker, but when you get letters closer together, you're going to find that they may get fuzzy on each other there.
Meanwhile, my Sharpie pen, which is my go-to even right here right now, liquid resin processes tend to have finer detail resolution because of the nature of the process. And the liquid is actually the base, and that gets photo cured instantaneously to harden in place into their shapes.
So you get less of that bleed because it's less intensely thermal on that process. It's not as super crisp and sharp as if I'm EDM and then molding afterwards, but still very, very good, best in class for additive manufacturing, and probably the highest detail resolution that you have with nice, clean edges.
FDM, I'm not knocking on it. FDM is fantastic if I want a 36-inch long part. FDM is the name of the game there. But it does sweat the small stuff. So on a layer by layer basis, because you are building in very concise layers and I'm depositing a filament from a real, essentially you think a line of material, I have broad, deliberate lines and zig-zagging back and forth on the interior to fill in that shape. So there also may be some micro gaps and things there. So imagine if I was drawing with a pencil, and I couldn't overlap. I couldn't overlap per layer. That's kind of the result that you get with a fill with FDM.
So we're going to move closer to kind of putting this together, but for me, it's really, really important to talk about these two subjects and understand that they have strengths, they have trade-offs, and these considerations because they're both really amazing technologies to explore. And spoiler alert, they're probably going to be tools in your tool belt. So depending on the project, there's going to be different lengths, different directions. But these are both extremely powerful tools for making tangible reality.
So let's just talk some pictures just because it's a nice little break to look at some pictures here really quickly but just really ingraining when I'm 3D printing. So these three pictures are 3D printed polymers. I'm able to have much more freedom in the design. You'll find that these designs tend to merit some of the goals that you have with injection molding, for example, even wall thicknesses, loving the hollowing-- or not necessarily hollowing, but lightweighting of parts, for example.
And I have a lot of options here. So SLA, I think we have over 15 unique SLA resins that you can choose from. I do have some fantastic thermoplastic plastics like Ultem here, which is just overall a superstar when it comes to ruggedization, durability, chemical resistance, and heat deflection.
And that's one that we can print up to 36 inches or so, although you can see it's a little bit more coarse than other processes, like for example, this HP Multijet Fusion part. And I'm showing one half here 3D printed with dyed black finish, and this one is dyed black, but it also has a vapor smooth post-processing finish which brings it closer to a molded look on that. But you can see different processes, again, they can exhibit different traits and different characteristics.
Going on the metals, I have a couple additive, and I have a machine part, and we'll just talk on that. But this is ExOne Binder Jet 420i, great composite metal. It's actually a blend of stainless steel and bronze because it's a secondary thermal process that infiltrates bronze so it's essentially 60% bronze, 40% steel here. But you can get more free-form geometries much more rapidly usually and more cost effective, but you can see that kind of matte finish and even some layer looks on the side here.
DMLS is actually more of a direct printing process for metal parts. It has more support structure considerations, actually a lot more support structure considerations to it. But again, you can see that really embraces organic shapes, lightweight shapes, designs that I may be designing for either die casting or metal injection molding. Usually a great start point for direct metal sintering.
But compare those matte natural finishes to a machine finish. So this part here is 6061-T6 aluminum CNC machined. You can see those mill marks here. So that would be where a cutting tool was spinning rapidly to create these features and remove material. This material has also been red anodized, which is why it's red.
And it has post-helical inserts installed to it as well, which sometimes some of you would do with a software metal so you have more repeatability if you're installing or removing it from an assembly, for example. But it is markedly smoother. The fact that I could actually run a process like anodized on it consistently and repeatedly is another thing there. So it's just something to consider is what am I doing. Not just what am I making, but what am I doing afterwards with that part?
And now let's talk about cost drivers. Actually, this is a slide that was really fun to make this, and I was trying to think about, what were the biggest cost drivers between these two because when it comes down to it, sometimes, yeah, I'm designing this part, I want to make it additive, I want to do this, and maybe I want to do additive metal. And you try to make it complex, and you end up getting this larger part that's a highly complex shape. And then you realize the pricing is still pretty significant.
So it has a lot of value add, but if you're trying to get to a larger, broader audience-- so instead of going into space with audience of 20, you're trying to get an audience of 20,000 or 200,000 or 2 million, it may not scale as much depending on these different cost drivers. And that's when you may go back and rethink, maybe, hey, should I simplify this, turn this into a different process that could scale to meet my customer demand here and the price point that they're willing to pay?
So here are some cost drivers to really talk about when I'm looking at choosing between CNC machining and 3D printing. So the first one is large geometries just tend to be easier to do with machining, and it has to do with high variability of different platforms across the additive manufacturing landscape as well as the fact that bulk is price driving in additive manufacturing. Adding material to my part is work for that machine in the additive manufacturing environment.
When I have a bulky geometry like a large plate with a few holes in it, in CNC machining, it's the opposite. It's getting that geometry there, cutting down the sides to get it to its net shape, and putting in the feature sets. But the less work it's doing, like less pocketing and lightweighting and stuff that's doing, the cheaper that job actually is. So usually, larger bulky geometries, machining gets a point there.
On the opposite, if I start coring this out, lightweighting the design, making it look more like an injection molded project regardless of the material that I'm choosing, additive starts to win a little bit more. Add lattice structures, add these things, they're inaccessible features so that you literally can't machine them, also an additive is definitely the design choice if you're going that direction.
Organic shapes, to say the same thing, down here-- I'm skipping one, sorry. I'll go back to off-angle holes here. But for organic features, it is the same thing. When I'm milling with a round tool, even a tool that has a round end to it, I have to do a lot of work and spend a lot of machine overhead in time to create organic shapes driving pricing way up, not to mention that I'm usually using a multi-axis machine like a fifth axis machine to not only tilt my tool head but also tilt my part to access these features. So I have a higher overhead machine. I have more machine time. I have higher pricing on my parts there.
Off-angle holes are just the beginning, right? So I'm going backwards one, but even an off-angle hole means I have to tilt that part. It's usually means that you'd need to add an extra operation for that, or I need to add a tilting axis like a fourth axis on my machine and run that part so I can hit that off-angle feature.
Because of that, 3D printing does have a great advantage. So if you do, for example, have a part with an array of mounting features that are all at different planes, if I'm multi-axis machining it, I need to recenter to that plane each time. But when I'm growing that part, it's just part of the part. It just grows layer by layer, and you're able to create those features pretty regularly.
Precision and tight tolerances, though, that is something where machining is still a winner, especially when you talk about metals and even like larger geometries. There's other thermal characteristics. There's other things that happen when you're additive manufacturing that could be a concern.
So even though a lot of our additive manufacturing products do have usually a plus or minus 0.005 tolerance for smaller features, the larger they get, the more you'll see that plus or minus 0.002 per inch, whichever is greater type caveat on the design and manufacturing standards for that process.
So when you need to hit position tolerances on a 3D printed part, usually you do your best. You look at our design guides, for example, Xometry guides, you work on your design, and you make that iteration one, but expect it to be iteration one. You measure. You see where it works. You see where there may be some fitment issues, and you look to actually adjust your CAD or model or adjust a feature to accommodate that for the next print.
So usually, there's a tuning process when I'm doing highly precise, repeatable parts in additive manufacturing, or I have to bring it to a CNC machine and do the post-machine operation, which essentially means I better be adding value some other way with that 3D printed part, for example, lightweight lattice structures or organic topology optimization.
Smooth surface finishes are inherent with CNC machined components. So obviously, when I'm milling these parts, I'm going to get at least a 125 RA, and I could always get that to a 63 or even better and bring it down to a smoother surface. And it's something I can do with the machining process itself. I don't need to do manual work afterward.
With 3D printing, the natural surface roughness of a part, especially the higher performance parts, tend to have some rougher finishes to them or be harder to post smooth afterwards. So it's usually a driver. It's usually a secondary process driver. We did just introduce chemical vapor smoothing to Xometry's SLS and MDF technologies.
Now that that auto quotes under the Finishing tab, it does significantly help there because it's a batch-based process so you're paying a smaller fee to do that, which is fantastic, but before that or for a lot of these processes, you're still doing a lot of manual labor so what you're actually paying is you're paying the person, not the automated process to do that type of smoothing. So that's why it is a cost driver for additive manufacturing.
And the last one I have on this list is deep holes and channels, especially inaccessible channels, right? So internal cooling channels on a device, for example, you just could not traditionally machine those, or you'd have to machine them in a very different way where you're drilling through, creating a intersect, and then plugging a couple of features in order to create a cooling channel, which is traditionally how it's done.
You could design that into your part. You can make larger holes, designs that may require typically a RAM or wire EDM in machining, which is a big cost driver within additive manufacturing. So it's definitely something to take advantage of, and it goes along with that complexity is free paradigm.
So recapping and putting this all together, here's some bullets that can really help you choose should I print or machine this part. So let's start with metal 3D printing. I'm looking for purpose-built designs. Is the value in the design? Because a lot of times, if it's a simple design, it's not cheaper to print it, and it's going to be cheaper, better, tighter tolerance to machine it. So what value are you adding? Am I integrating features, creating complex pieces within an all-in-one assembly like what Dixon Valve was doing? Is that what I'm working on? If not, then yeah, maybe you look at some other processes there.
This is also fantastic for cast surrogates. So we're talking about machining versus printing here, but I've seen a lot of cast parts, so investment cast metal parts, move into a just-in-time production with metal 3D printing because you just don't have the setups, the tooling, and the weeks of wait time, especially for part demands that are lower dozens. So single digits, lower dozens per year are really great candidates to move to additive manufacturing technology.
Polymer 3D printing, much more robust, much more proliferated in the world from anywhere from one-offs to fixtures to fit checks. Smaller pieces actually can be very competitive with molding. A small piece, I've done some tests before. If it's about the size of my thumb or less, I can probably build about 2,000 to 2,500 in the additive manufacturing process before I do a break even with injection mold tooling. So really powerful there. And the larger part is, the more that break even moves backwards in the lower quantities where a part this big may actually be 50, for example.
And something to understand is that we are in a world of internet of things. We're in a world of high design iterations. If you're planning on tooling up for a process or you're planning to machine this part out, but you know that your PC board may change, it may be worth doing your first rounds with additive manufacturing or find a way where additive manufacturing is a component that's surrounding that part that may have a rev change.
It's really cool because you can mix these technologies. And I know I used to work at a place where we did just that where we had molded out our shells, and essentially, we had a generic mounting plate made using SLS nylon for our PCDs. So if certain things like ICs or capacitors, resistors changed the design a little bit, we didn't scrap our tool. We just changed the digital file in our 3D printing method.
But going to CNC, why is CNC still super popular and amazing and will 3D printing destroy CNC? No, they're just going to complement each other, right, because CNC still does larger parts and higher production super well. Critical requirements, characterized tolerances, the materials itself, stringent specifications are everything that machining, it's already there. It's just part of the process now.
And it's something really to note that when I am working on a high-end product like an aerospace component, for example, with additive, it may require a mixture of engineering. So it's not just a print to go experience. It's sometimes engineering, iteration, testing, simulation. There's other work that goes into it beyond just making. Making's the last thing that you do with additive on super critical products where machining, it's like does it make the print? And if you're able to do that, provide specification documents required, you're typically very good to go.
So then you have folks like BMW who sit back and say, why not both? BMW tooling engineers actually use Xometry because we are a consolidated supply chain for them. These jigs that you see on the left, these are hand-held tools to actually install decals. I actually think these ones actually install the BMW logo on a vehicle in their assembly plant.
But this is a mixture. They have something that's covering a contoured surface, and that main body there is white PC through FDM. So it is FDM polycarbonate material, and they take advantage of this handheld fixture in being able to do these organic shapes kind of free-form design using 3D printing for those main unibody designs. They also take advantage of sparser infills that you could do with FDM in particular to make the parts lighter weight to decrease essentially fatigue in using these parts.
But surrounding this, they have features that need to indicate very precisely to the vehicle components themselves. So these are installing trim. I need to actually touch point on parts of the vehicle that are consistent on that vehicle in order to make sure that those decals are spot on every single time. They want that material to be rigid. They want it to be highly precise. And that's where you see CNC machined white Delrin as well as CNC aluminum alloy on these fixtures.
So BMW has really embraced, first off, the power that you can do where you can just order these online like you're ordering from Amazon, but they're custom components through Xometry, but also, embrace different technologies really highlighting the strengths of each one of those there.
And that's something just really important to note is that 3D printing, it does bridge product development and production. It's a fuzzy line now because you can introduce a product with additive manufacturing components and still move to other processes down the line. It's successful. It's very, very flexible.
When I'm looking at machined components, I may be doing initial checks with a 3D printed surrogate, for example. I got a part that's ultimately designed for fifth axis machining here, and this part is great to make it for PLA for just very cheaply to get something very quick in hand and validate my design, do some testing as I'm designing for that machine component. So these things just really, really go hand in hand.
And this is where I kind of want just to talk through some of the tools that we have at Xometry because we only have 45 minutes, an hour to talk today, but I really want to talk about these resources because you can actually use Xometry's tools, including free CAD add-ins, our instant quoting site to compare those costs. Click, drag, look at CNC machine on pricing. That's great. Move to DMLS or move to binder [INAUDIBLE] metal. You can take a look at that and look at those materials and compare cost price, lead times, and get feedback as well on that.
It's a really powerful tool, especially if you're looking to move forward with a project. It's something that you can just go ahead and use regularly because it's a free site to use. And it is just very, very useful, especially if you're at the design stage where as you change or try out different things, what happens if I combine these features together? You can see the price consequence on the tail end there.
The other thing to note is that we do a lot of resources because not everybody knows a dozen manufacturing technologies back and forth. So if you go to xometry.com/resources, it's going to take you to the main page there, but definitely check out our videos on our YouTube channel. I work on a lot of our design guides. So we have a design guide for every single process that we offer at Xometry.
And we're always highlighting customers' experiences and putting some very interesting stuff on our blog. So definitely please go ahead, sign up for our blog, and take a look. But we're here to help. We're here to continue the conversation. And I know I love manufacturing. I love making parts. So if there's any way, shape, or form that we could help each other out, that's fantastic.
So again, thank you so much. Xometry's a global company so that's why the .com and eu because we have a European division as well. But always happy to connect and speak more about choosing between additive manufacturing, CNC machining, or other processes. Thanks so much for your time.