Latest / Hardware Startups with Fexingo: Devices, Robotics, and Manufacturing Tech Companies / How Hardware Startups Use Ultrasonic Welding to Join Plastics
Transcript
- Lucas: If you've ever wondered how those sleek plastic enclosures on smart rings and wireless earbuds stay sealed without a single visible screw, the answer is often a process called ultrasonic welding. Luna: I've seen videos of it — two plastic parts vibrate against each other, and they just fuse together. It looks almost like magic. Lucas: It's not magic; it's 20 kilohertz of mechanical vibration applied under pressure. The energy creates friction at the interface, generating enough heat to melt a thin layer of plastic, which then solidifies into a molecular bond. Luna: And it happens in under a second, right? That's why it's so appealing for hardware startups trying to scale assembly. Lucas: Exactly. Compared to using adhesives — which have cure times, outgassing, and application consistency issues — or screws, which add parts and labor, ultrasonic welding is clean and fast. No consumables, no chemicals, just electricity and the parts themselves. Luna: But it's not for every plastic. I assume you need the right material. Lucas: Right — it works best with amorphous thermoplastics like ABS, polycarbonate, and acrylic. Semi-crystalline plastics like polyethylene are trickier because they have a sharper melting point and don't transmit vibrations as well. Most consumer electronics enclosures are ABS or a pc abs blend, so it lines up well. Luna: What about the equipment? I imagine a startup can't just buy a $50,000 welder on a whim. Lucas: That's actually where it gets interesting for early-stage companies. You can rent time on a machine at a prototyping shop for maybe $50 to $100 an hour. And if you're doing a small run — say a few hundred units — you can manually feed parts. The capital expense really hits when you need automation for thousands of units a day. Luna: So a startup could validate the design with a bench-top welder, then invest in a production cell once they have traction. Lucas: Exactly. And the tooling — the horn that contacts the part — is often the bigger cost. A custom horn can run $500 to $2,000, and you need one for each different joint geometry. But if you design your enclosure so that the weld line is the same across multiple products, you can reuse the same horn. Luna: That's a design for manufacturing lesson: standardize your joint geometry from the start. Lucas: Right. And the joint itself needs what's called an energy director — a small triangular ridge on one of the parts that concentrates the vibration energy at the tip. That ridge melts first and flows into the joint, creating a consistent bond. If you don't have an energy director, you get inconsistent welds. Luna: So the mold design has to account for that. It's not just a flat mating surface. Lucas: No, it's a tiny feature — often 0.3 to 0.5 millimeters tall — but it's critical. I spoke with a smart-ring startup that switched from screws to ultrasonic welding in their second production run. They had to tweak their injection mold to add the energy director, which cost about $3,000 in modifications. But they eliminated four screws per device and cut assembly time by 80 percent. Luna: Eighty percent? That's huge. What was the volume? Lucas: They were doing about 10,000 units initially. With screws, each unit took about 45 seconds of manual assembly. With ultrasonic welding, it dropped to under 10 seconds per unit, including loading and unloading. And they achieved IP67 waterproofing without a gasket. Luna: No gasket means fewer parts, lower BOM cost, and fewer failure modes. Lucas: Exactly. The weld itself can be hermetic if designed correctly. For wearables that need to survive sweat and rain, that's a big win. The trade-off is that the joint is permanent — you can't open the device for repair. But for something like a smart ring or a disposable medical sensor, that's often acceptable. Luna: That makes me think about repairability — there's a tension between sealing a device and allowing the user to replace a battery. Lucas: That's a real consideration. Some companies use ultrasonic welding for permanent seals and then design a separate access panel held by screws for the battery. Or they use a weaker weld that can be broken open with a tool, though that risks damaging the enclosure. Luna: One thing I've wondered: does the vibration damage sensitive electronics inside the enclosure? Like a tiny accelerometer or a microphone. Lucas: It's a legitimate concern. The frequencies used are typically 20 to 40 kilohertz — well above human hearing, but they can resonate with internal components. In practice, the vibration amplitude is small — usually 10 to 50 micrometers — and the duration is short, under a second. Most surface-mount components can survive it. But if you have a MEMS sensor with a fragile membrane, you might need to pot it or add a damper. Good design practice is to run a vibration analysis or just test a few units. Luna: So prototyping is key. You can't just assume it'll work. Lucas: Absolutely. And there's another factor: the horn must contact the part precisely. If the horn shape doesn't match the part contour, you get uneven pressure and weak welds. That's why horn design is often iterative. Startups that do this well typically run several trials with a contract manufacturer that specializes in ultrasonic welding before committing to production tooling. Luna: And I'm guessing the same principle applies to the anvil — the bottom fixture that holds the other half. Lucas: Right. The anvil supports the lower part and must be rigid. If it flexes, you lose energy and get a cold weld — where the parts touch but don't bond. That's a common failure mode for first-time users. They set the pressure too low or the anvil isn't stiff enough. Luna: Speaking of first-time users, what's the learning curve for a startup team? Can a mechanical engineer pick it up in a day? Lucas: The basic operation is straightforward: you set the pressure, amplitude, and weld time. But optimizing those parameters for a new part geometry takes trial and error. A good contract manufacturer can help dial it in. I know a startup that sent their enclosure to a welding shop and the shop had optimized parameters within two hours. That's the kind of resource you want. Luna: It also helps that ultrasonic welding doesn't require ventilation for fumes, unlike solvent bonding or some adhesives. Lucas: That's a big plus for lean operations. No curing oven, no chemical storage, no hazardous waste disposal. The energy consumption is also low — a typical welder draws maybe 1,500 watts during the weld cycle, and the cycle is under a second. So the overall energy cost per part is negligible. Luna: Let's talk about a concrete example. You mentioned the smart-ring startup. What was their material? Lucas: They used a medical-grade ABS, which is common for wearable devices. The enclosure had two halves that needed to be sealed after the electronics were inserted. They had been using four small screws — that's four extra parts to buy, four extra holes to mold, and a gasket to keep water out. The screws also added weight and required a secondary operation to install. Luna: And the gasket had to be compressed evenly. That's another tolerance stack-up. Lucas: Exactly. With ultrasonic welding, they eliminated the gasket, eliminated the screws, and reduced the number of mold actions needed. The cost savings on the BOM alone was about 15 cents per unit — which at 10,000 units is $1,500. But the bigger saving was in labor: they went from a two-person assembly station doing 80 units per hour to one person doing 360 units per hour. Luna: That's a 4.5x improvement in labor productivity. For a startup, that can mean the difference between profitable and not. Lucas: Precisely. And the welding equipment — they bought a used bench-top welder for $4,000. The horn and anvil cost another $1,500. Total capital outlay was $5,500 for the tooling. They recouped that in labor savings within about 1,500 units. After that, every unit was cheaper. Luna: That's a classic hardware ROI story. And it's not just for wearables — I've seen ultrasonic welding used in medical devices, automotive sensors, even some consumer drone frames. Lucas: Right. Anywhere you need a clean, strong, permanent joint in plastic. The key is designing for the process from the start. If you try to retrofit ultrasonic welding onto an existing screw-boss design, you'll struggle. You have to design the joint line, the energy director, and the wall thickness around the welding parameters. Luna: So for a hardware startup listening, the takeaway is: if you're designing a plastic enclosure, consider ultrasonic welding early. It might let you skip fasteners, reduce part count, and seal your device for free. Lucas: If these conversations are useful for what you're building or running, it's worth mentioning that this show stays independent thanks to listeners who chip in a couple of dollars a month. That genuinely covers hosting costs and makes sure we don't have to run ads or chase sponsors. Luna: Yeah, it's a small thing that makes a real difference. If you've gotten something out of these episodes, buy me a coffee dot com slash fexingo — every bit helps keep the podcast going. Lucas: So to build on that, ultrasonic welding isn't the only option. There's also laser welding, which is even more precise but more expensive, and vibration welding, which works for larger parts. For a startup, ultrasonic is usually the sweet spot. Luna: And what about snap-fits? Those are common in consumer products. Lucas: Snap-fits are great for design for assembly — no tools, no consumables — but they have limitations. They can loosen over time, they don't provide a hermetic seal, and they require more complex mold features. Ultrasonic welding gives you a permanent bond and can seal simultaneously. For many startups, that trade-off is worth it. Luna: I remember reading about a startup that tried snap-fits on a waterproof speaker and had field failures after six months. They switched to ultrasonic welding and solved it. Lucas: That's a classic case. The snap-fit loses its retention force as the plastic creeps under constant load. Ultrasonic welding doesn't have that issue because the bond is molecular. Once cooled, it's as strong as the base material. Luna: So for products with a long expected life — like a medical device that might be used for years — ultrasonic welding is a more reliable choice. Lucas: Exactly. And it's also worth noting that ultrasonic welding can join dissimilar plastics, as long as they're compatible. ABS to polycarbonate works well, for example. That gives designers flexibility. Luna: But there are limits — you can't weld a thermoplastic to a thermoset or to metal without special techniques. Lucas: Right. And the joint must be accessible to the horn. If your enclosure has deep internal cavities or a complex 3D shape, you may need multiple welds or a custom horn. That's where cost can escalate. Luna: So it's not a silver bullet. It's a tool for a specific set of problems. Lucas: Exactly. But for the problems it solves — fast, clean, permanent plastic joining — it's hard to beat. And with the rise of low-cost benchtop welders from companies like Sonics and Branson, it's more accessible to startups than ever. Luna: So the next time I see a seamless plastic device, I'll assume it was welded at 20 kilohertz. Lucas: Good bet. And now you know what to ask your contract manufacturer about if you're designing your own enclosure.