Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / Hardware Startups Use Fused Deposition Modeling for Bridge Tooling
Transcript
- Lucas: If you're a hardware startup waiting eight weeks for a steel injection mold, you are burning cash. You have no product in the field, no user feedback, no revenue. So the smart ones don't wait — they print their tooling. Luna: Print tooling? Like, a 3D printer spits out an injection mold? Lucas: Exactly. Fused deposition modeling — FDM — on a high-temperature printer. You print the mold cavity and core out of something like Ultem 9085 or PPSU, mount it in a standard mold frame, and shoot production-grade resin into it. It's called bridge tooling. Luna: And the trade-off? Plastic tools don't last as long as steel. Lucas: Right. You get maybe a few thousand shots before the cavity wears or the gate cracks. But for a startup doing initial production runs of 500 to 3,000 units, that's enough. And you get parts in two weeks instead of eight. Luna: So what's the real-world case? I remember reading about a medical device startup that did exactly this. Lucas: Yeah, I was thinking of the same one. A company developing a single-use surgical stapler — needed 2,000 units for a clinical study. Steel mold lead time was nine weeks. They printed FDM inserts in eleven days for twelve hundred dollars. Shot all 2,000 parts over two weekends. The steel mold arrived three weeks after the study started. Luna: So they had parts in hand and were enrolling patients while the toolmaker was still cutting steel. That's the whole argument. Lucas: That is the whole argument. And the material choice matters. You cannot use standard PLA for an injection mold — it'll deform under pressure within ten shots. You need a high-heat thermoplastic. Ultem 9085 is the gold standard because it can handle mold temperatures around 150 degrees Celsius and injection pressures up to maybe 60 bar. But it's expensive — about three hundred dollars a kilogram spool. Luna: There's also the surface finish difference. A printed tool leaves layer lines on the part. Lucas: Yes. So for cosmetic parts, you either post-machine the cavity or accept a textured finish. For functional parts that get painted or covered, it's often fine. The medical stapler example — the parts were inside the handle assembly, never seen by the surgeon. So layer lines didn't matter. Luna: Where does FDM bridge tooling fall down? I've heard cycle times are longer. Lucas: Cycle time can be thirty to fifty percent longer than with a steel mold because the printed tool doesn't conduct heat as well. You have to wait longer for the part to cool before ejection. The medical stapler's cycle was about forty-five seconds versus twenty-eight seconds on steel. But when you only need two thousand parts, the slower cycle is still faster than waiting for steel. Luna: And the cost comparison — twelve hundred dollars for the printed inserts versus, what, twenty thousand for the steel mold? Lucas: Roughly. The steel mold was eighteen thousand, but it'll last for hundreds of thousands of shots. The bridge tooling cost one thousand three hundred bucks including material and setup time. So for a run of two thousand, the per-part tooling cost is sixty-five cents versus nine dollars with steel. But if they run a hundred thousand units later, steel wins. Luna: Another example I like — a consumer robotics company that made a small cleaning robot. They used FDM bridge tooling for the first three production batches while they finalized the design. Lucas: Oh, I know that one. They iterated the housing geometry three times based on early user feedback. Each time they printed a new insert in four days instead of revising a steel mold over three weeks. Total tooling cost for three design spins was about four thousand dollars. If they'd cut steel each time, it would have been over fifty thousand. Luna: And the part cost was higher per unit because of the longer cycle and shorter tool life. But the speed to market was six months faster. Lucas: Exactly. And that's the metric that matters for early-stage hardware — time to market, time to feedback. The per-part cost penalty is trivial compared to the cost of delay. Luna: Speaking of cost — these conversations are exactly why we keep the show ad free. A couple of dollars a month from listeners who find them useful is genuinely what keeps these going. Buy me a coffee dot com slash fexingo, if you've gotten something out of them. Lucas: Yeah. No pressure, but it adds up. We don't run ads, we don't do sponsorships — it's just us and the listener. That direct support makes a real difference. Luna: Alright. Back to the printed tools. What about the geometry limits? You can't do complex slides or lifters, right? Lucas: That's the big constraint. FDM bridge tooling is best for relatively simple two-part molds — a cavity and a core. If your part requires side actions, unscrewing mechanisms, or multi-stage ejection, printed tooling gets very difficult. You can sometimes design around it, but at that point you might be better off with a machined aluminum insert, which is a different trade-off. Luna: Aluminum is faster than steel but slower than FDM. And more expensive than printed, but longer-lasting. Lucas: Right. Aluminum bridge tooling is a middle ground — about six weeks lead time, maybe five thousand dollars, and it can run ten to twenty thousand shots. FDM is two weeks and under two thousand. So if your initial run is under five hundred parts and you need them yesterday, FDM wins. If you're doing five thousand and might scale, aluminum is safer. Luna: What about the printer itself? Not every startup has a high-temperature FDM machine. Lucas: You outsource it. Companies like Protolabs, Xometry, and Stratasys Direct offer FDM bridge tooling as a service. You upload your CAD, they quote you in hours. The medical startup used Protolabs. The robotics company used a local service bureau. The key is specifying that the part is a mold insert — they'll orient the print and select the infill pattern for compressive strength. Luna: Infill matters. You need high density — like ninety percent or more — so the cavity doesn't collapse under pressure. Lucas: Exactly. And you want a solid shell, maybe three or four perimeters. The service bureaus know this. But if you try to print it yourself on a desktop machine with low infill, you'll get ten parts before the cavity cracks. Luna: So the playbook is: design for a simple two-plate mold, use a high-temp material like Ultem, print at high infill, and accept the surface finish you get. Lucas: That's it. And plan for the tool to be disposable. You're not building a production asset — you're building a bridge to get your first parts into the hands of real users. The steel mold is still coming. But by the time it arrives, you're already iterating based on feedback. Luna: I've also seen startups use FDM bridge tooling for pre-production runs for trade shows. They need fifty demo units for CES — steel mold won't be ready in time. Lucas: Classic use case. A startup we covered last season did exactly that — printed inserts, shot fifty parts in a weekend, assembled the demos, and had a booth with working product. The steel mold arrived a week after CES. They used the show feedback to modify the design before cutting production tooling. Luna: It's almost a no-brainer for any hardware startup with a timeline under six months to first production. Lucas: I'd go further — if you have a timeline under twelve months, you should at least evaluate it. The cost of the printed tool is so low that even if you only use it for fifty parts, it's cheaper than waiting. And you learn something about your design's moldability before you commit to steel. Luna: That's a good point. The printed tool itself is a test of the part geometry. If the part doesn't fill properly in FDM tooling, it probably won't fill in steel either — but you find out for two thousand dollars instead of twenty thousand. Lucas: Exactly. And you can modify the gate location or add venting on the next print iteration overnight. That's the real advantage — speed of learning. Steel molds punish mistakes. Printed molds forgive them. Luna: Alright, so for any hardware startup listening: if you're facing an eight-week mold lead time and you only need a few thousand parts, look into FDM bridge tooling. It might save you a quarter. Lucas: And if you try it, let us know how it goes. We'd love to hear real numbers on tool life and cycle time. That's the kind of data that helps the whole community.