What a 3D printing patent actually covers
Nike holds an additive manufacturing 3D printing patent for footwear. This page explains the process it describes, where the same mechanics break down, and how to decide between printing and CNC machining for a real part. Written for engineers who have to pick a process, not read a press release.
What the additive process in the patent does
Nike's 3D printing patent describes a way to build part of a shoe directly on fabric. A 3D printer deposits polymer and resin onto a textile layer held flat on the build plate, then the deposit is cured into a solid pattern. The result is a printed structure bonded to cloth in one setup. No adhesive, no stitching step in between.
This matters because it removes an assembly operation. In a conventional shoe build, the sole, the upper, and any reinforcement are made separately and joined later. Every joint is a possible failure point. Printing the lattice straight onto the textile means the bond is formed while the material is still reactive, so the interface is chemical as well as mechanical.
The patent is not about a printer model. It is about the sequence: flatten the substrate, print the pattern, cure in place. That sequence is what makes the process repeatable enough to file. For an engineer, the interesting part is the constraint it implies. You need a flat, dimensionally stable substrate, and you need a print head that can survive being centimetres from a curing lamp.
The same sequence shows up outside footwear. Printing a gasket onto a stamped flange, a wear pad onto a metal plate, or a grip pattern onto a handle all follow the same logic. The patent is narrow. The principle is not.
Where the 3D printing patent process stops working
Additive wins when the geometry is complex and the load is light. It loses when you need isotropic strength, tight tolerances, or a surface that has to seal. Printed polymer is layered, so it is anisotropic by construction. A part printed flat can be strong in X and Y and weak in Z. Design around the layer direction or the part will delaminate under load.
Tolerance is the second wall. Most industrial polymer printers hold ±0.1 mm to ±0.3 mm on a good day. That is fine for a grip or a cover. It is not fine for a bearing bore, a mating spigot, or anything that has to slide. If your drawing says ±0.05 mm, printing is the wrong first step.
Surface finish is the third. As-printed surfaces carry layer lines. You can bead blast, tumble, or vapor smooth, but each of those adds a step and can change dimensions. If the part needs Ra 0.8–1.6 μm and a specific flatness, plan on machining or printing oversize and finishing.
Cost flips with volume, too. Printing has almost no tooling cost, so one unit is cheap and 10,000 units are not. A machined part has setup cost up front and a low marginal cost after that. The crossover is usually somewhere in the low hundreds for simple parts and much higher for complex ones.
Material choices and what they mean for the part
The patent language covers polymer and resin, which is a wide net. In practice, the choice comes down to load, temperature, and chemical exposure. PLA and ABS are cheap and easy but creep under sustained load. PC and PA are tougher and hold threads better. PEEK and PEI survive heat and solvents but need a hot chamber to print well.
For parts that carry structural load, filled materials help. Carbon fibre reinforced nylon raises stiffness and reduces creep, at the cost of a rougher surface and faster nozzle wear. It also makes the anisotropy worse in the Z direction, so the same design rule applies: orient the layers along the load path.
Metal printing exists, but it is a different conversation. It needs support removal, stress relief, and often a finish pass. For a bracket or a housing, machining from 6061 or 17-4PH is usually faster and cheaper unless the geometry is genuinely impossible to cut.
If the part touches skin, food, or a medical device, check the resin. Some photopolymers are not biocompatible and not stable under UV. That is a materials qualification question, not a printer setting.
A practical route from print to finished part
Most production parts do not stay printed or machined. They move between the two. A common route is to print a near-net shape and then machine the critical features. That gets you the lattice or the internal channel from printing and the bore, the face, and the thread from cutting.
The order matters. Print oversize by 0.3–0.5 mm on any face that will be cut, then fixture on a stable printed surface. Printed polymer can be fixtured in soft jaws, but it deflects. Take light passes and check the first part before running the rest.
For prototype runs, printing is the fast path to a fit check. Once the geometry is frozen, move the load-bearing parts to 5-axis machining. On a 4,000 × 400 × 150 mm travel machine, you can hold ±0.005 mm across a part that no printer would touch.
If you need both, say so at quoting. A supplier who runs additive and subtractive under one roof can print the blank and cut the critical faces without shipping the part twice. That saves a week and a re-fixturing error.
3D printing vs CNC machining: which one fits
Use this as a first filter. If a row points both ways, the tighter tolerance wins.
| Requirement | 3D printing | CNC machining | Pick |
|---|---|---|---|
| Tolerance | ±0.1–0.3 mm typical | ±0.005 mm achievable | CNC |
| Unit cost at qty 1 | Low, no tooling | Setup cost applies | 3D printing |
| Unit cost at qty 10,000 | High, slow per part | Low marginal cost | CNC |
| Internal lattice or hollow | Native, no tooling | Requires access or split | 3D printing |
| Isotropic strength | Layer dependent | Uniform from billet | CNC |
| Surface finish | Layer lines, needs work | Ra 0.8–1.6 μm as cut | CNC |
| Metal parts | Limited alloys, post work | Wide alloy range | CNC |
| Design still changing | Edit the file, reprint | Re-cut or re-fixture | 3D printing |
The verdict
If the geometry is complex, the load is light, and the design is still moving, print it. If the part has a bore, a seal face, a thread, or a tolerance under ±0.05 mm, machine it. For parts that need both, print the blank and machine the critical features in one shop.
Questions engineers ask next
Does a 3D printing patent stop me from printing my own part?
Patents cover specific claims, not the general idea of printing. Nike's 3D printing patent covers a particular sequence of printing onto a textile substrate and curing in place. Printing a bracket, a jig, or a housing does not touch those claims.
If your process copies a claimed step-for-step sequence for a commercial product, get a patent attorney to read the claims. For general prototyping and production tooling, this is not a practical concern.
How tight a tolerance can additive hold before I need machining?
Industrial polymer printers typically hold ±0.1 mm to ±0.3 mm, and that is on a stable geometry with good thermal control. Feature size, orientation, and support marks all push the number around.
Below ±0.05 mm, plan on a machining pass. Print oversize and cut the critical features. That is standard practice and usually cheaper than chasing printer settings.
Can printed parts replace machined metal parts?
Sometimes, if the load is low and the environment is mild. Printed polymer creeps under sustained load and softens with heat, so a printed bracket that holds at room temperature may sag at 60 °C.
For structural parts, machined aluminium or stainless is still the default. Use printing for covers, grips, ducts, and prototypes, then move to metal once the design is fixed.
What surface finish should I expect from a printed part?
As-printed surfaces show layer lines. The roughness depends on layer height, usually 0.1–0.3 mm, and on the print angle. Vertical walls look better than shallow overhangs.
Bead blasting, tumbling, or vapor smoothing can improve the look, but each adds handling and can shift dimensions. If the drawing calls for Ra 0.8–1.6 μm, machine the face instead of printing and finishing it.
When does machining become cheaper than printing?
It depends on part complexity more than quantity. For a simple part, machining wins somewhere in the low hundreds of units because setup cost is spread thin. For a complex part with internal channels, printing can stay cheaper much longer.
The honest answer is to quote both. A shop that runs additive and subtractive can give you both numbers in one pass, which is faster than guessing.
Can I get both processes from one supplier?
Yes, and it usually saves time. GreatLight runs 3D printing alongside 127 high-precision CNC machines, including 16 simultaneous 5-axis centers. A printed blank can go straight to a machining cell without a second shipping step.
Send the drawing and tell us which features are critical. We quote both routes in the same 12-hour window and mark the faces that need cutting.
Send the drawing, get both routes
Tell us which features are critical and we will quote printing and machining side by side. Quotation and free DFM analysis within 12 hours, production can start within 24 hours.
12-hour quote100% inspectionNo MOQNDA on request
Follow GreatLight
We publish setup notes, tooling trials and inspection data from the factory floor.
