Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Cut Costs with Multi-Cavity Molds
Transcript
- Lucas: There's this moment every hardware founder hits — you've got a working prototype, maybe a few hundred units out in the field, and your injection mold is a single-cavity tool running a forty-second cycle. Luna: And you realize you're going to die of old age before you hit ten thousand units. Lucas: Exactly. So the obvious solution is a multi-cavity mold — instead of one part per cycle, you get four, eight, sixteen parts. But the decision isn't that simple, because the tooling cost scales faster than linearly, and the engineering gets a lot more demanding. Luna: Let's start with the economics. What's the typical jump in tooling cost from one cavity to, say, eight? Lucas: Rough rule of thumb: doubling the cavity count roughly doubles the mold cost, but not quite — maybe 1.6 to 1.8 times. So a single-cavity mold might be $15,000, an eight-cavity might be $80,000 to $100,000. The per-unit savings come from cycle time. With one cavity you get one part per cycle. With eight, you get eight parts in maybe only twenty percent more cycle time, because the injection and cooling steps overlap. Luna: So your output per hour goes up by a factor of six or seven, but your tooling cost only went up by a factor of six. That starts to pay back fast. Lucas: It does, but the break-even quantity depends on your part complexity and material. I talked to a startup that makes smart water sensors — the plastic housing is about the size of a deck of cards. They started with a single-cavity prototype mold, running ten thousand units a year. Their per-part cost from the molder was $1.20. When they moved to a sixteen-cavity production mold, the per-part cost dropped to seventeen cents. Luna: Seventeen cents — that's a seven-to-one reduction. How many units did they need to run to justify the new mold? Lucas: The sixteen-cavity mold cost them about $120,000. The single-cavity mold was $18,000. So the delta is roughly $100,000. At a savings of $1.03 per part, they break even at around 97,000 units. They were projecting 150,000 units over two years, so it was a no-brainer. Luna: And they didn't have to wait for the big mold. They kept running the single-cavity tool for initial production while the multi-cavity was being built — that's a smart dual-track approach. Lucas: Exactly. You don't bet the farm on tooling until you've validated demand. But here's the engineering catch: when you go multi-cavity, you have to ensure each cavity fills identically. If one cavity fills faster than the others, you get short shots in the slow cavities and flash in the fast ones. That's called gate balance. Luna: And it's not just the gate, right? Cooling has to be balanced too. If one cavity runs hotter, the part shrinks differently. Lucas: Right. Cooling channel design becomes critical. In a single-cavity mold you can get away with a simple straight-through cooling line. In a multi-cavity, you need baffles, bubblers, or conformal cooling channels to keep every cavity within a few degrees of each other. Otherwise you get warped parts and inconsistent dimensions. Luna: What about the runner system? That gets more complex too — you're feeding many cavities from one sprue. Lucas: Yeah, you have two main options: cold runner or hot runner. Cold runner is simpler and cheaper — the plastic solidifies in the runners, and you regrind them. But the waste adds up. Hot runner systems keep the plastic molten in the manifold, so there's zero runner waste. They also give better control over gate timing, which helps with balance. Luna: But hot runners are more expensive and require more maintenance, right? I've heard horror stories about hot runner systems clogging. Lucas: They can. If you're running a material that degrades easily — like some flame-retardant plastics — the residue can build up in the manifold. But for high-volume production, the cost savings on material usually outweigh the maintenance. For the water sensor startup, they went with a hot runner because they were running a standard polycarbonate blend, and the runner scrap would have been about fifteen percent of their material cost. Luna: Let's talk about when NOT to go multi-cavity. If your total volume is only fifty thousand units over the product's life, is a four-cavity mold worth it? Lucas: Probably not. The rule of thumb I use is: if your total run is under 100,000 units, stick with a single or two-cavity mold. Above that, start looking at four or eight cavities. But it also depends on part size. For tiny parts like connectors or medical device components, you can justify sixteen or thirty-two cavities even at lower volumes because the tooling cost per cavity is smaller. Luna: And there's the lead time factor. Multi-cavity molds take longer to build — maybe ten to sixteen weeks versus six to eight for single-cavity. If you need to get to market fast, a single-cavity or family mold might get you there sooner. Lucas: That's a great point. A family mold — where you have different cavities for different parts of the same product in one tool — can be a good intermediate step. You get multiple parts per cycle, but the cavities are different shapes, so balancing is even harder. But it saves you from buying multiple molds. Luna: I want to circle back to something you said earlier about the water sensor startup. Their unit cost went from $1.20 to seventeen cents. That's not just about the mold — it's also about cycle time savings and labor. Did they see any other operational benefits? Lucas: Yeah, a big one is consistency. With a single-cavity mold, every part comes from the same cavity, so you have zero cavity to cavity variation. But in multi-cavity, you have to qualify each cavity separately. The best practice is to run a capability study on each cavity — measure critical dimensions on a sample of parts from each cavity and make sure all cavities are within spec. If one cavity is off, you might have to rework the mold. Luna: And if you're running a six-sigma process, that qualification is essential. But for a startup just trying to ship product, it can feel like overhead. Lucas: It is overhead, but it's the kind that prevents a recall. I've seen a startup skip cavity qualification, and they ended up with one cavity producing parts that were three percent larger than the others. They didn't catch it until field failures started coming in. Luna: Ouch. So what's your practical advice for a founder listening who's about to pull the trigger on a multi-cavity mold? Lucas: First, get a mold flow analysis done. It's not expensive — a few thousand dollars — and it'll tell you if your gate locations and cooling channels are balanced. Second, talk to at least three mold makers and get their recommendations on cavity count. Some will push for more cavities because it's more money for them. Others will be conservative. Listen to the one who asks detailed questions about your part geometry and material. Luna: And ask for references from other hardware startups. A mold maker who mainly works with automotive companies might not understand your need for speed and flexibility. Lucas: Absolutely. One last thing: consider a mold with interchangeable cavity inserts. That way, if you need to add a cavity or replace a damaged one, you don't have to scrap the entire mold. It costs more upfront, but it gives you flexibility as your design evolves. Luna: That's a great tip. Before we wrap, I want to mention something practical for our listeners. If you found this episode useful — and I hope you did — these shows stay ad-free thanks to listener support. You can help keep it that way by visiting buy me a coffee dot com slash fexingo. Lucas: Yeah, it's a simple way to say 'this was worth my time.' And we really appreciate it. No pressure, just if the show has saved you from a costly mistake, that's what the coffee is for. Luna: Exactly. So back to the topic — one more question: what's the biggest mistake you see hardware startups make when they spec a multi-cavity mold? Lucas: They underestimate the cooling time. They assume that because they're getting more parts per cycle, they can run the mold faster. But if the cooling is not balanced, they have to run at the speed of the slowest cooling cavity. So the actual cycle time might be longer than expected, and the per-unit cost doesn't drop as much. Luna: So the lesson is: invest in the cooling design upfront, or you'll pay for it in cycle time later. Lucas: Exactly. And if you can, run a design of experiments on your mold — try different injection speeds, pack pressures, and cooling times to find the sweet spot. That's how you get the most out of your multi-cavity investment. Luna: Alright, Lucas. For our listeners who are scaling production, I hope this gives you a clear framework. Next episode, we're going to talk about something that's been on my mind: how to choose between ultrasonic welding and snap-fit assembly for your product enclosure. Lucas: That's a good one. See you next time.