Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / Why Hardware Startups Are Switching to Metal Binder Jetting
Transcript
- Lucas: You know, when I talk to hardware founders about metal 3D printing, they almost always default to powder bed fusion — the laser sintering approach. But there's another process that's quietly eating its lunch in production speed and cost, especially for medium-run parts. Luna: You're talking about metal binder jetting. I've seen it called 'the inkjet for metal parts.' Lucas: Exactly. And it's not new — the technology has been around since the 1990s at MIT. But in the last three years, the machines have gotten reliable enough that startups are using them for end-use production, not just prototypes. A robotics company I spoke with recently switched from CNC machining a complex bracket to binder jetting and cut their per-part cost by 40 percent. Luna: Forty percent is hard to ignore. But what's the catch? There's always a catch with new manufacturing processes. Lucas: The big one is post-processing. With binder jetting, you print a 'green part' — metal powder held together by a polymer binder. It's fragile, like a chalky aspirin. Then you have to debind it in a furnace and sinter it to full density. That sintering step causes shrinkage — typically 15 to 20 percent linear — and it's not always perfectly uniform. Luna: So you have to design the green part bigger and hope the shrink is predictable. That sounds like a headache for tight tolerances. Lucas: It is, but the machine manufacturers have gotten much better at modeling it. HP's Metal Jet system, for example, uses a combination of simulation and closed-loop sintering controls to hit tolerances within plus or minus 0.2 percent on most geometries. That's competitive with investment casting. Luna: And build speed? I've heard binder jetting is dramatically faster because the print head doesn't have to melt the powder — it just deposits binder. Lucas: Right. In powder bed fusion, a 200-watt laser has to trace every contour of every layer, melting the metal. That's slow. In binder jetting, the print head sweeps across the entire powder bed in one pass — it's like printing a page of text versus handwriting each letter. The same robotics startup I mentioned went from a six-week lead time to five days for a batch of 200 brackets. Luna: Five days from six weeks. That changes how you think about inventory and supply chain risk. Lucas: Exactly. And it's not just lead time — the lack of support structures is a huge advantage. In powder bed fusion, every overhang needs a metal support that you have to cut off later, which adds labor and material waste. Binder jetting packs the entire build volume with loose powder, so unsupported features are fine. You can print a hollow sphere with internal channels and it just works. Luna: That opens up designs you couldn't machine or cast. But I want to push back on one thing: material properties. Sintered parts are inherently porous, right? They don't reach full density like a wrought material. Lucas: Historically, yes. But modern binder jetting with hot isostatic pressing — HIP — can close the porosity and achieve over 99.5 percent density. For stainless steel 316L, that yields tensile strength and elongation comparable to annealed bar stock. So for structural applications that aren't fatigue-critical, it's absolutely viable. Luna: What about materials beyond stainless? Can you print copper, titanium, tool steel? Lucas: Copper is actually one of the best use cases because it's difficult to laser-sinter — copper reflects infrared light. Binder jetting handles it easily. Tool steels like H13 are also common. Titanium is trickier because it's reactive during sintering, but a few companies offer Ti-6Al-4V with good results. The material library is expanding fast. Luna: So what should a hardware founder consider when deciding between binder jetting and other metal additive processes? Let's say they're making a small production run of a complex part. Lucas: First, volume. For fewer than 50 parts, powder bed fusion might still be cheaper because the setup cost for binder jetting — including the sintering furnace cycle — has a fixed overhead. Above 100 parts, binder jetting wins on both cost and speed. Second, geometry: if your part has internal channels, lattice structures, or requires no support removal, binder jetting is ideal. Third, material: if you need a high-performance alloy like Inconel, powder bed fusion has a more mature supply chain. Luna: And what about surface finish? Binder jetting leaves a matte, slightly rough surface compared to machined finishes. Lucas: That's true. As-sintered surface roughness is typically around 6 to 10 microns Ra. If you need a better finish, you're looking at secondary operations like tumbling or machining critical surfaces. But for many structural or internal components, that roughness is acceptable. Luna: I've also heard that the binder itself can leave carbon residue after debinding. Does that affect corrosion resistance? Lucas: It can, especially in 316L. The carbon can sensitize the stainless steel if not properly controlled. But the leading binder systems now use water-soluble or low-residue polymers. And the sintering atmosphere — typically argon or vacuum — helps manage carbon pickup. It's a known variable that experienced suppliers have dialed in. Luna: Let's talk economics. What's the machine cost range for a binder jetting system that a startup could actually afford? Lucas: Entry-level systems from companies like Desktop Metal or ExOne — now part of Desktop Metal — start around $150,000 to $200,000. But that's just the printer. You also need a debinding system and a sintering furnace, which can add another $100,000 to $150,000. So the total investment is roughly $300,000 to $400,000 for a turnkey setup. That's significant, but it's within reach for a well-funded hardware startup, especially if they plan to run it in-house. Luna: Or they can use a service bureau. There are now several that specialize in binder jetting, like Mantle or 3DEO. That lowers the barrier to entry. Lucas: Absolutely. And that's probably the smartest first step — validate your design, run a few batches, and only buy the machine once you have consistent demand. The service bureaus have the process tuned, and they can give you cost estimates that help you model the business case. Luna: I want to step back and think about the bigger picture. Binder jetting seems like a classic example of a technology that was under the radar for years and is suddenly hitting an inflection point. What changed? Lucas: Three things. First, print head reliability. The industrial piezo inkjet heads from Fujifilm and Konica Minolta now have lifetimes of billions of droplets. Second, software — the ability to simulate and compensate for sintering shrinkage with predictive models. Third, a whole ecosystem of post-processing equipment that's become standardized. Five years ago, you had to engineer your own debinding and sintering recipes. Now it's plug and play. Luna: That's exactly the kind of maturation that lets startups adopt a technology without having to become experts in it. Lucas: Right. And if these conversations are useful for what you're building or running, I'll mention that the show stays ad-free thanks to listener support. If you find value in episodes like this, you can support us at buy me a coffee dot com slash fexingo. Every little bit helps keep these deep dives coming. Luna: Yeah, and it's a small way to give back if you're getting practical takeaways. Appreciate everyone who has chipped in. Lucas: Alright, back to binder jetting. One more angle I want to hit: there's a specific class of part where binder jetting is almost a no-brainer, and that's large, thin-walled enclosures. Think of an electronics housing for a drone or a robotic arm controller. Those are typically made from sheet metal or plastic, but if you need EMI shielding or thermal conductivity, metal binder jetting can produce them in one piece with integrated mounting bosses and heat sink fins. Luna: And you avoid the assembly and welding of sheet metal. That's a real cost saver. Lucas: Exactly. I saw an example from a startup making industrial sensors: they had an aluminum enclosure that was previously a five-piece assembly — machined base, sheet metal cover, plastic cable gland, the works. They redesigned it as a single binder-jetted part in 316L. Total part count went from 12 to 3, and assembly time dropped from 20 minutes to 2. Luna: That's the kind of metric that makes a CFO sit up and listen. But what about design for binder jetting? Are there rules of thumb? Lucas: The big one is to avoid unsupported horizontal holes that are too small — they can fill with powder and not clean out. But that's manageable. Another is to keep wall thickness above 0.5 millimeters for green part handling. And because there are no thermal stresses during printing, you can design with very thin walls and sharp corners that would warp in powder bed fusion. Luna: So it's almost the opposite design constraints. More freedom in some ways, but you have to think about the entire process chain from green to sintered. Lucas: Exactly. And that's where the real learning curve is — not in the printing, but in understanding how the part shrinks and changes shape. The good news is that simulation tools are improving fast. I've seen startups use simple FEA to predict distortion and then iterate the green part geometry in a day or two. Luna: What's on the horizon? I've heard about continuous binder jetting — like a conveyor belt system. Lucas: Yes. Desktop Metal and others are developing systems where the build box moves continuously, so you can have multiple part runs in one machine without stopping. That could push throughput even higher. And there's research into multi-material binder jetting — printing with different powders in the same layer to create graded properties or embedded sensors. Luna: That sounds like it could be a game changer for smart hardware products. Lucas: It's early stage, but the potential is real. For now, the practical takeaway for any hardware founder is: if you're producing metal parts in volumes of 100 to 10,000 units per year, and your geometries are complex, binder jetting is worth a serious look. The cost per part at those volumes often beats CNC machining and investment casting. Luna: And you get the design freedom of additive without the speed penalty of laser melting. It's a different tool for a different job. Lucas: Right. And knowing which tool to use is half the battle. Next episode, we'll look at how hardware startups are combining binder jetting with conventional machining in hybrid workflows — that's where things get really interesting.