CNC Processing of Glasses: How Frame Parts Are Machined

• 1
  What CNC processing of glasses actually removes
• 2
  Material behavior decides the cutting strategy
• 3
  Curved eyewires and hinge pockets need more than three axes
• 4
  Where tolerance matters and where it does not
• 5
  Surface finish changes feel, reflection, and wear
• 6
  Which process fits which frame part

A glasses frame is a set of small, thin, curved parts joined at tight tolerances. The front carries two lens apertures and a bridge; the temples pivot at hinges and fold flat against the front. Most of these parts start as solid stock. A machined front is cut from a plate or block, not bent from sheet. The cutter removes material along three or five axes until the curved eyewire and the bridge emerge as one piece.

That distinction matters. Stamped or molded frames reach shape by forcing material into a die. CNC reaches shape by subtraction, so the geometry lives in the tool path, not in a mold. You can change the bridge width, the eyewire bevel, or the hinge pocket by editing the program. There is no tooling cost per revision. For small runs and frequent design changes, that is the main reason engineers pick cnc processing of glasses over injection molding.

The trade-off is cycle time. Cutting a full acetate front takes minutes of spindle time, while molding takes seconds. Subtractive work also wastes stock. A 12 mm acetate block may become a 6 mm front, with the rest turned into chips. On thin stainless temples the loss is small; on thick acetate fronts it is large enough to show up in unit cost at high volume.

Where the process wins is in the parts that must hold size: hinge plates, lens grooves, screw bosses, and the folding stops that keep temples from over-rotating. These are the features a wearer feels every day, and they are the features that a loose mold struggles to hold.

• 1
  Shape from tool pathNo mold, so design edits cost programming time only.
• 2
  Small runs stay viableOne prototype and 10,000 parts use the same setup.
• 3
  Watch stock lossThick acetate fronts convert much of the block to chips.

Acetate is the classic frame plastic. It cuts cleanly at high spindle speed and leaves a polished edge when the feed is low and the cutter is sharp. Heat is the enemy. If the tool rubs instead of slicing, the acetate smears, and the eyewire groove fills with soft material. Keep the chip load up and the spindle fast. A roughing pass followed by a light finishing pass gives a better edge than one deep cut.

Titanium is the other common answer. Grades like TA2 and TC4 (Ti-6Al-4V) give frames that are light and spring back instead of bending. They also cut poorly compared with aluminum. Titanium conducts heat away from the cut slowly, so the tool tip runs hot. Use sharp carbide, generous coolant, and moderate surface speed. Rushing titanium temples is how you get chatter, burned edges, and tools that fail early.

Stainless grades 304, 316L, and 17-4PH appear in hinges, screws, and nose pad arms. They machine more predictably than titanium but still work-harden if the cutter dwells. 17-4PH can be machined in the solution-treated state and aged afterward, which keeps the cutting soft and the final part hard.

For temples that must feel warm and flex, some designs use carbon fiber or reinforced polymer. These cut as abrasive composites. Tool life drops, dust control matters, and the finish is usually left as-machined rather than polished.

• 1
  AcetateHigh speed, light finish pass, control heat.
• 2
  TitaniumSharp carbide, coolant, moderate speed.
• 3
  StainlessAvoid dwelling so the cut does not work-harden.

A flat front can be cut on a three-axis machine. The cutter moves in X, Y, and Z while the part stays still. That works when the eyewire curve is gentle and the hinge pockets open straight down. Many rimless and semi-rimless fronts fall in this group, and three-axis work keeps the setup simple and the cost low.

Once the frame wraps around the face, the geometry turns three-dimensional. The lens plane tilts inward, the bridge sits on a compound curve, and the hinge pocket faces sideways. A three-axis cutter can reach some of this by tilting the part in a fixture, but each angle needs a new setup. Every setup adds a chance for position error.

Five-axis machining solves this by moving the tool and the part together. The spindle tilts while the table rotates, so the cutter stays normal to the surface through the whole eyewire. That keeps the groove width consistent and avoids the faceting you get when a ball cutter approaches a curve from one angle. Hinge pockets cut this way come out square to the folding axis, which is what makes the temple open smoothly.

Four-axis work sits between the two. It adds rotation around one axis, usually enough for temples with a single sweep and for fronts with a simple wrap. If the part needs undercuts on two sides, five axes is the practical choice.

• 1
  Three-axisFlat fronts, straight hinge pockets, simplest setup.
• 2
  Four-axisSingle-sweep temples, mild wrap fronts.
• 3
  Five-axisWrapped fronts, compound curves, two-sided undercuts.

Not every surface on a frame needs the same accuracy. The lens groove is the critical one. If the groove width drifts, the lens sits loose or will not seat. Hold the groove to a tight band and check it against the lens edge, not just against the drawing. A groove that measures right but does not grip the lens is still wrong.

Hinge alignment is the second critical zone. The two hinge halves must line up so the screw passes through without force and the temple folds without binding. The screw hole diameter and the pocket depth both feed into this. A few hundredths of a millimeter off, and the temple droops or rubs. This is where cnc processing of glasses earns its keep, because the same program repeats the position on every part.

Bridge width and overall frame width are cosmetic-functional. They affect fit, but a small drift does not stop the frame from working. Temples are similar. A temple that is 0.1 mm thinner than nominal still folds and still sits on the ear. Spending five-axis time to chase a tenth on those surfaces usually is not worth it.

The practical rule: put the tight tolerance on the interfaces, and leave the free surfaces at a looser band. That keeps cycle time and scrap down. Our general machining tolerance is ±0.005 mm (±0.0002 in), but we only apply it to features that need it.

• 1
  TightLens groove, hinge screw holes, pocket depth.
• 2
  LooseBridge width, temple thickness, outer profile.

Finish is not decoration on a frame. It controls how the surface feels against skin, how light reflects, and how the part resists wear. A temple that is too rough will drag on hair. A front that is too polished will show every fingerprint. Machining leaves its own texture, and the finish step builds from there.

As-machined surfaces land around Ra 1.6–3.2 μm. That is a fine matte look and a common choice for titanium temples where the raw metal is the design. A finer cut brings it to Ra 0.8–1.6 μm, which reads as satin. Bead blasting over a machined surface gives an even, non-reflective skin that hides tool marks and small scratches from handling.

Polished acetate and metal use tumbling and buffing to reach Ra 0.2–0.8 μm. This is a slow step because polishing removes material unevenly, so it should be planned after the critical dimensions are already in spec. If you polish first and then chase a tolerance, you cut through the polished layer.

Coating comes last. Anodizing adds color and a hard oxide layer on aluminum frames. Plating covers stainless and brass hinge parts. Laser marking handles logos and size text, with a minimum character height of 1.5 mm so the mark stays legible after finishing.

• 1
  Ra 1.6–3.2 μmAs-machined matte, common on titanium.
• 2
  Ra 0.8–1.6 μmSatin, good for front surfaces.
• 3
  Ra 0.2–0.8 μmPolished, plan after dimensions are set.