Polished Nickel Mirror: How CNC Machining Decides the Final Finish
A polished nickel mirror is not made by polishing alone. The surface you see at the end is mostly decided by the cutting path, the tool load and the heat that went into the part hours earlier. This page explains the mechanism, the boundaries, and how to judge whether your part geometry suits 5-axis machining or needs a different route.

In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
What a Polished Nickel Mirror Actually Is
A polished nickel mirror is a metal reflector: a nickel surface flat enough that light reflects without scattering. Unlike a glass mirror, the reflective layer is the structural part itself. Light hits nickel, and nickel reflects it. There is no silvered backing to peel.
That changes the specification. You are not buying a coating. You are buying a machined surface whose flatness, waviness and roughness all matter at the same time. A scratch 2 μm deep breaks the reflection. A 1 μm wave across 50 mm bends it.
Nickel earns its place here for three reasons. It resists corrosion without a protective layer on top. It takes a high polish because it is ductile enough to flow under a polishing pad rather than fracture. And it holds that finish in humid or mildly acidic service where aluminium would cloud over.
Typical parts are small and awkward: laser cavity end plates, beam-steering mirrors for optical benches, spectrometer fold mirrors, mirror mounts with cooling channels behind the reflective face. The reflective face is often only 20–80 mm across, but the mounting features around it decide whether the face stays flat.
- 1RoughnessRa 0.2–0.8 μm is the working band for optical reflection
- 2FlatnessSet by the cut and the fixturing, not by the polishing pad
- 3SubstrateNickel plating over copper, brass or steel is common
Why Nickel Punishes a Bad Cutting Strategy
Nickel is ductile and work hardens fast. The first pass cuts cleanly. Every pass after that meets a surface that has already been strained and hardened by the tool that just left. Cutting forces climb, and the tool starts pushing material instead of shearing it.
When that happens the damage is not a visible gouge. It is a layer of deformed metal, torn micro-pits and built-up edge fragments smeared across the face. Polishing removes the top few micrometres, but the tears underneath stay. They show up later as haze or as a reflection that never quite comes sharp.
Heat makes it worse. Nickel conducts heat poorly compared with aluminium, so cutting heat concentrates at the edge. Above roughly 300–400 °C at the interface, adhesion between chip and tool grows quickly. The chip welds to the cutting edge, breaks off, and leaves a mark.
The fix is mechanical, not chemical. Keep the radial depth of cut light, keep the feed per tooth high enough to cut rather than rub, and never let the tool dwell. A cut that rubs is the single most common cause of a mirror face that will not polish out.
Sharp, uncoated carbide or a polished-flute tool for nickel alloys works better than a general-purpose coated insert. A coating that flakes contributes hard particles straight into the surface you are trying to protect.
- 1Light radial stepoverKeeps the tool engaged in shear instead of rubbing
- 2High feed per toothThins the deformed layer under the cut
- 3No dwellTool pauses create weld marks that polishing cannot remove
What 5-Axis Machining Changes for a Polished Nickel Mirror
On a three-axis machine the tool axis is fixed. On a curved or angled mirror face, the effective cutting speed changes constantly across the surface, and the tool contact point drifts toward the tip where speed drops to zero. That is where smearing starts.
A five-axis machine tilts the tool so the contact point stays in a controlled position. The cutting speed stays even across the whole face, and the scallop pattern becomes regular and shallow instead of patchy. Regular scallops polish out predictably. Patchy ones do not.
The second gain is setup. A mirror with features on several faces normally needs multiple setups, and every re-clamp introduces a small error and a new chance to distort a thin part. Cutting the reflective face and its mounting datum in one setup keeps them related to each other.
The third gain is access. Undercut edges, curved rims and angled mounting bosses can be reached without long, flexible toolholders that chatter. Chatter on a mirror face leaves a periodic wave that no amount of hand polishing removes.
This is why the polished nickel mirror workflow at our shop starts on the five-axis side. We run 16 simultaneous five-axis machining centers alongside 12 four-axis mills and 27 three-axis machines, so the geometry picks the machine rather than the other way round.
- 1Even contact speedTilting keeps the tool out of the zero-speed tip zone
- 2Fewer setupsFace and datum cut in one clamping
- 3Shorter toolsLess overhang means less chatter on thin faces
When This Route Is the Wrong Choice
Five-axis nickel machining is not the cheapest way to get a reflective surface, and it is not always the right one. Knowing the boundary saves money and time.
If the part is flat, small and produced in the thousands, a plated flat lapped after machining will beat a machined mirror face on both cost and flatness. Nickel plating a lapped steel substrate gives you a very flat reflector without any five-axis time at all.
If the mirror only needs to look reflective and no one measures wavefront, a bright decorative finish is enough. Bead blasting followed by polishing gets there for a fraction of the cost. Optical flatness is a different purchase from shine.
If the part is larger than 4,000 mm in any axis, it is outside our travel. If it is a thin nickel shell under 0.8 mm, clamping force alone will distort it and the polishing step will chase a shape that keeps moving.
And if the design calls for a mirror face on a part with deep internal channels, expect to split the part or accept a compromise. You cannot reach the bottom of a deep pocket with a tool stiff enough to leave an optical surface.
- 1High volume flat partsLapped and plated substrates usually win
- 2Decorative onlyPolish alone is enough; skip the optical spec
- 3Thin shellsClamping distortion outruns any polishing gain
How the Machined Surface Becomes a Mirror
The machined face leaves the machine at Ra 1.6–3.2 μm as-machined, or Ra 0.8–1.6 μm when the finishing pass is run with a small stepover and a fresh tool. That is the starting point, not the finish line.
Polishing then removes material in stages. Coarse stages cut the scallop peaks down. Fine stages reduce roughness without cutting new scratches. If the peaks are regular, each stage removes a predictable amount. If the surface has torn metal and weld marks, the polishing pad digs into the soft spots and the face goes wavy.
That is the whole argument in one sentence: polishing amplifies whatever the cutter left. A good cut polishes quickly and evenly. A bad cut polishes unevenly and never quite settles.
We target Ra 0.2–0.8 μm on finished optical faces. Getting there depends on the cut more than on the pad. A part that arrives at polishing with a clean, uniform scallop pattern typically needs fewer stages and holds flatness better through them.
Tolerance control runs alongside this. We hold ±0.005 mm on critical features, and the reflective face is measured against its own mounting datum, not against an arbitrary edge that may itself be out of position.
- 1As-machinedRa 1.6–3.2 μm baseline from a standard finishing pass
- 2Fine machinedRa 0.8–1.6 μm with small stepover and a sharp tool
- 3PolishedRa 0.2–0.8 μm target for reflective faces
What We Control Before the Part Reaches Polishing
The reflective face is usually a small feature on a part with a lot of other work in it. That is where process control matters more than any single operation.
We check raw material before cutting. Nickel and nickel-plated stock varies in hardness between lots, and a lot that is 15% harder will change the cutting behaviour enough to matter on a mirror face. In-process monitoring catches the drift before the finishing pass.
Every part is inspected before shipment, and reports are available on request. For a mirror face, the inspection we care about is the flatness and roughness of the reflective surface relative to its datum, plus a visual check under oblique light for tears and weld marks.
Fixturing gets designed around the reflective face, not around the easiest clamp. Soft jaws, vacuum chucks and low-pressure supports keep a thin mirror from bowing while it is cut.
Our plants in Dongguan and Singapore run 127 high-precision CNC machines across 7,600 m² with 150 technicians. The five-axis capacity is what makes the single-setup approach practical at production volume.
- 1Incoming materialHardness variation between lots is checked first
- 2In-processSurface condition monitored before the finishing pass
- 3FinalFlatness, roughness and oblique-light visual check
Choosing the Right Route for a Reflective Nickel Part
Match the part to the process before you ask for a quote.
| Part condition | Best route | Why |
|---|---|---|
| Curved or angled mirror face | 5-axis machining | Contact speed stays even; no tip smearing |
| Flat face, 5,000+ pieces | Lap and plate | Machining time per part is too high |
| Multi-face mirror housing | 5-axis, one setup | Face and datum stay related |
| Decorative shine only | Bead blast then polish | Optical flatness is not being measured |
| Thin nickel shell under 0.8 mm | Redesign or support | Clamping distorts the part |
| Part over 4,000 mm | Outside our travel | Split the design or change supplier |
Which Route to Pick
If your mirror face is curved, angled or tied to a datum on another face, machine it on a 5-axis center in one setup. If it is flat, small and running in the thousands, lap and plate a substrate instead and skip the machining time entirely.
Questions Engineers Ask Next
Can you machine a mirror face and then plate it with nickel?
Yes, and it is a common route. We machine the substrate to the required geometry, then nickel plating builds the reflective layer. The plated layer follows the machined surface, so the cut quality still controls the final flatness.
If plating thickness varies across a large face, flatness can drift. On parts where that matters, we machine after plating instead and polish the nickel directly.
How flat can a machined nickel face be before polishing?
It depends on the part more than the machine. A rigid, well-supported face in the 20–80 mm range holds flatness well within our ±0.005 mm tolerance on critical features. Thin unsupported faces move under clamping and cutting loads.
We usually recommend leaving enough wall thickness so the face can be supported during the finishing pass, then trimming the support material afterwards if the design allows it.
Does polishing change the dimensions?
Yes. Polishing removes material, and on a mirror face the amount is not negligible. We plan an allowance into the finishing pass so the polished face lands on the nominal dimension.
Tell us at the quoting stage that the face will be polished. If we machine to the final dimension and polishing happens afterwards, the part will come in undersize.
What nickel alloys do you machine?
We machine nickel plating over copper, brass and steel substrates, plus nickel alloys in the titanium and special material group such as Inconel. Copper alloys C101, C103, C110 and beryllium copper are common substrates for mirror work.
If your drawing names a specific alloy, send it with the quote request and we will confirm machinability and finishing route before cutting.
What is the minimum order quantity for a machined mirror part?
There is no minimum order quantity. We run from one prototype to 10,000+ part runs on the same equipment.
For a single prototype we would typically machine the reflective face and inspect it before deciding on the polishing stages, so the first part tells us what the process needs.
How do you protect the finished face during shipping?
Polished faces are covered with a removable protective film and packed so the reflective surface does not contact anything rigid. Parts are packed individually when the face is large or the flatness is tight.
If your assembly process needs the film to stay on until the last step, say so on the order and we will mark the packing accordingly.
Send Us the Drawing Before You Commit to a Route
Upload your part and we will return a quotation with a free DFM analysis within 12 hours, including a note on whether the mirror face should be machined or lapped.
12-hour quoteFree DFM analysis100% inspectionNDA on request