Detroit Lion 3D Printed Helmet: How Metal Additive Parts Get Made

• 1
  What a Detroit Lion 3D printed helmet actually demonstrates
• 2
  Layer by layer: how metal additive builds the part
• 3
  Alloys that make sense for printed metal hardware
• 4
  Where printing stops and CNC finishing starts
• 5
  When printing beats machining, and when it does not
• 6
  Printed metal blank vs billet machining: quick comparison

Padded football helmets have used 3D printed lattice liners since 2023. A Detroit Lion 3D printed helmet matters to engineers because it put a printed structure inside a life-safety product, where every gram and every load path gets inspected. The public discussion stops at weight and fit. The useful part is the mechanism underneath.

Lattice geometry replaces solid foam or solid metal. Under impact, a lattice crushes progressively: struts bend, then buckle, then densify. That sequence sets a force plateau instead of a sharp spike. Foam does something similar, but foam is stochastic. Its cell size varies from batch to batch, so its crush curve varies too. A printed lattice is deterministic. Change the strut diameter and you move the plateau on purpose.

That is why sports equipment and medical implants converge on the same idea. Both need a structure that absorbs energy at a controlled rate in a small envelope. Both need stiffness where the load attaches and compliance where the body meets the part. A printed lattice lets one part do both jobs.

The manufacturing consequence is less glamorous. A lattice is only as good as its smallest feature. Struts under roughly 0.4 mm are hard to build cleanly in metal and hard to inspect. If your design calls for finer detail than your printer can hold, the crush curve you simulated will not be the crush curve you ship.

Metal laser powder bed fusion, usually called DMLS or SLM, spreads a thin layer of metal powder, melts the cross-section with a laser, then drops the build plate and repeats. Layers run 20–60 μm in most production work. A 40 mm tall lattice can be 700 to 2,000 layers.

Each melt pool cools fast, on the order of 10⁵ to 10⁶ K/s. That speed produces a fine cellular microstructure, which is often stronger than the same alloy cast. It also produces residual stress. Thin walls and long unsupported spans warp as the part cools, so support structures carry both the overhangs and part of the thermal load.

Supports are the hidden cost. They must be strong enough to hold the part flat and weak enough to break off without gouging the surface. On a lattice part, internal supports may be impossible to reach at all, which is why lattice orientation on the build plate gets decided before the geometry is frozen.

Post-processing is where the tolerance lives. As-built surfaces land around Ra 8–12 μm and rarely beat ±0.1 mm on critical features. Anything that must seal, slide, or mate needs machining. That is a CNC operation, not a printing operation.

Titanium Ti-6Al-4V (TC4) is the default for load-bearing printed parts. High specific strength, good corrosion resistance, and a mature print recipe. It machines slowly and wears tools, so keep machined features simple and let the print do the complex geometry.

17-4PH stainless prints well and can be aged to higher strength. It is a practical choice when you need corrosion resistance plus a hardness number, and it machines far more easily than titanium. 316L is the workhorse for corrosion and cleanability, but it is soft and will not hold a wear surface.

Aluminium alloys such as AlSi10Mg print readily and give low weight, though they are not as strong as wrought 7075. Inconel grades handle heat and are common in aerospace ducts and combustor parts. They are also the most expensive to print and to finish.

We machine all of these in the same shop. If your part is mostly printed with a few tight bores, we print the blank and cut the critical features on a 5-axis center. If the geometry is mostly prismatic, printing is usually the wrong answer and billet machining wins.

Treat the printed part as a near-net blank. As-built metal additive holds roughly ±0.1 mm on well-supported features and worse on thin walls and long spans. That is fine for brackets, ducting, and lattice cores. It is not fine for bearing bores, seal faces, or anything with a mating fit.

Machining the blank brings features to ±0.005 mm (±0.0002 in) and surface finish to Ra 0.2–0.8 μm where needed. Typical callouts land at Ra 0.8–1.6 μm for sealing faces and Ra 1.6–3.2 μm for general mating surfaces.

The catch is setup. A printed blank has no flat reference and no straight edge. We usually print a stock boss or pad that becomes the datum, then hold the part in a fixture machined to match. Without that datum, the first cut is guesswork.

Add machining stock on purpose. Leave 0.3–0.5 mm on faces that will be cut, and 0.5 mm or more on bores. Printing to net size and hoping to skim 0.05 mm off usually ends with a scrapped part and a reprint.

Printing wins when the geometry is internal, organic, or consolidated. A lattice that absorbs energy, a conformal cooling channel inside a mold insert, a manifold that replaces twelve welded tubes. These shapes are expensive or impossible to cut from billet.

Printing also wins on low-volume consolidation. If a bracket assembly currently needs five machined parts, eight fasteners, and two weld operations, one printed part can remove the joints. Fewer joints means fewer failure points and less assembly labor.

Printing loses on simple prismatic parts. A plate with holes, a shaft, a housing with square pockets. Billet machining is faster, cheaper, and holds tighter tolerance on those shapes. Printing them adds cost and heat treatment without adding function.

Printing also loses on large solid sections. A 100 mm cube of solid metal is slow to print, expensive per kilogram, and carries heavy residual stress. If most of your part is solid, start from bar stock and cut it.