How Future Cars Could be Designed and Made Like Smartphones

Nov 15, 2023

Related Video

Kevin Czinger reckons cars can be designed and made like smartphones, with brands developing gotta-have-it products with help from artificially intelligent and machine-learned supercomputers, with construction farmed out to a deglobalized constellation of contract manufacturers. Additive manufacturing, or 3D printing, is key to his vision and is the mission of his company, Divergent 3D.

When I first covered Czinger's idea in 2017, his vehicle architecture involved 3D-printed aluminum nodes connecting commodity carbon-fiber tubes cut to length. Today the carbon tubes are out, and the nodes have evolved into vastly smarter structures like subframes (for the Aston Martin DBR22), suspension components, a transmission case (by Xtrac), and perhaps soon an engine block.

The Czinger 21C hypercar we've ridden in is a glamorous proof of concept for a broader vision where low-investment, low-carbon manufacturing enables affordable transportation. Among the most impressive developments made since 2017 are the computer-aided design algorithms Czinger's firm has developed, which leverage Bidirectional Evolutionary Structural Optimization (BESO). This iterative design process determines not only the shape of the component to be produced by the 12-laser, powdered-metal sintering "printer" but also the ideal alloy it's made of.

Engineers start by defining a part or structure's basic geometric shape, its mounting locations, the loads it's expected to bear under normal conditions, the deformation it must allow in a crash, fatigue requirements, repairability concerns, environmental operating conditions (e.g., will it be exposed to corrosive elements or extreme temperatures?), and finally weight and cost targets. Once all of these conditions are programmed in, Czinger's algorithms select an ideal metal alloy for the part, choosing from among 28 affordable, generally accessible elements. It also defines the ideal shape, placing material only where necessary for structure, fluid flow, or other reasons. Finally it programs the most laser-energy efficient 3D-printing layer strategy for the part.

Where necessary, the computer can subdivide a structure or part into smaller pieces that fit within the given 3D printer, recommending the best approach for joining individual pieces—either by adhesive bonding or (for repairable parts) mechanical fastening. Bonded parts get the tabs and slots or windows necessary for applying and/or curing the various adhesives. One of these oven-cures in 60 minutes, and another Czinger-patented one cures in 2.0 seconds using ultraviolet light. The latter is used to "tack weld" parts headed for the oven. The goal is for every part to emerge from these forming and bonding processes requiring no further heat treatment or corrosion countermeasures and only minimal machining and post-processing prior to final assembly.

The result is a light, strong part that reduces the total parts count and eliminates subassembly processes, like a Tesla gigacasting but without the giga tooling costs of purchasing a gigantic casting rig and forming complex dies for multiple parts. Vastly less energy is required to produce each piece, and product design changes are as simple as flashing new code to the printer. There's also minimal disincentive to switching from producing one part to producing another in the same machine.

Czinger's proposed assembly line occupies a hexagonal footprint about 75 feet across that can assemble 10,000 rolling chassis or 100,000 subframes per year. This line takes two to three weeks to set up and commission, about like the 3D printing machines, making this small-footprint manufacturing operation very quick to scale and extremely adaptable.

With such an operation, the break-even point for a given product virtually disappears, and the consequences of a failed product design are nearly eliminated because with minimal traditional tooling, the machinery can be programmed quickly to make parts for another car or another industry. Warehousing bulky service parts becomes unnecessary, as replacements can be printed on demand.

I'm most enthralled with Czinger's 3D-printed engine block concept. He's not ready to talk specifics, but imagine routing all the coolant and oil in passages through the block in the best manner without regard for loading in or shaking out sand cores. Oil passages and squirters could be placed almost anywhere, coolant flow could be directed in ways never before possible, and the engine might not need any external fluid plumbing—heck, maybe the alloy could even be altered locally to suit different parts of the engine. Final engine machining and assembly would be vastly reduced, reliability improved, and the exterior appearance is bound to be way cooler.

When I lament the pending demise of combustion engines, Czinger counters with an impassioned argument in favor of e-fuels produced from recycled CO2—a cause 3D-printed engines give me fresh impetus to champion.