Divergent 3D Printed Car: What the $100M Hexagon Bet Actually Buys

• 1
  Why the Divergent 3D printed car is a joint problem, not a printing problem
• 2
  How a printed node becomes a structural part
• 3
  Where additive plus machining beats a conventional chassis
• 4
  Boundary conditions that decide whether this route fits
• 5
  Printed node versus machined node versus welded tube cluster

The Divergent 3D printed car gets attention for its printed metal nodes. The nodes are the small, load-bearing connectors where tubes meet. Printing gets you a topology that you cannot mill out of one block: hollow internal ribs, graded wall thickness, and organic transitions between struts. That is real, but it is half the story.

The other half is the joint. Each node arrives from the printer near net shape, then gets machined where it matters: the bonding faces, the tube sockets, and any bolt circle. Adhesive bonds the node to the carbon or metal tube. The joint, not the node, sets the stiffness of the whole structure.

That split explains the Hexagon investment. Hexagon sells metrology, CAD/CAM, and inspection software. A printed node is only useful if you can measure it, compare it to the model, and decide whether the bond gap is right. Software and sensors are the bottleneck, not the laser.

So when you read about the $100 million, read it as money for the digital thread around the print. Design, simulation, machining of bond faces, and metrology of the assembled joint. Remove any one of those and the car does not come together.

Metal printing here means laser powder bed fusion, usually in an aluminum alloy such as AlSi10Mg or a titanium grade. Layer thickness sits around 30–60 μm. After the build, the node is cut off the plate, stress relieved, and often heat treated. Support removal follows.

Then the part goes to a machining center. The bond faces need flatness and a controlled surface. On a 5-axis machine we hold ±0.005 mm on critical features and Ra 0.8–1.6 μm on sealing or bonding surfaces where the spec allows. Tighter finishes, Ra 0.2–0.8 μm, are possible when a joint needs it.

Print tolerance is looser than machining tolerance. That is the whole reason the hybrid route works. You let the printer make the hard shape, then you let the cutter make the accurate face. The two processes cover each other's weak spots.

Assembly comes last. Adhesive is applied, the tube is inserted, and the joint is fixtured until cure. Gap control is the critical variable. Too thick and the bond is soft, too thin and the adhesive starves. That is why the machined socket diameter matters more than the printed node's outer contour.

Low volume is the first condition. A stamped steel or cast aluminum chassis needs tooling. Tooling needs volume to amortize. At a few hundred vehicles a year, the tooling cost per car dominates. Printing nodes and bonding tubes skip that tooling entirely.

Second, geometry that carries load in more than one direction. A node can merge five or six tube directions into one continuous load path. A welded tube cluster needs gussets and still leaves stress concentrations at the welds. A printed node spreads that stress over a larger area.

Third, iteration speed. Change a tube length and you change one cut tube. Change a node and you reprint it. No die, no fixture redesign at the stamping level. For a program that is still tuning stiffness and crash behavior, that matters.

None of this means the printer replaces the machine shop. Tubes are cut and sometimes turned. Inserts, bushings, and suspension pickup points are often machined from 6061-T6, 7075, or 17-4PH. Those parts see the point loads.

Cost per kilogram is the hard limit. Printed aluminum is far more expensive per kilo than extruded tube or castings. The route only pays when the part count is low or the geometry cannot be made another way. High-volume commuter cars will not switch to printed nodes.

Build envelope is the second limit. Powder bed printers have a size ceiling, and the bigger the part, the worse the residual stress and the higher the scrap risk. Nodes stay small for a reason. Long structural members stay as tubes.

Qualification is the third. Bonded joints are hard to inspect. You cannot see through the adhesive line. Programs rely on process control, coupon testing, and sometimes destructive teardown of witness parts. That is slower to certify than a welded or bolted joint, and it drives testing cost.

Fatigue data is thinner than for welded steel. Adhesive bonds behave well in shear and poorly in peel. Design the joint so the load path stays in shear. If the node sees peel, the joint will not last, no matter how good the print is.