Precise CNC alloy processing: what changes when the material changes
Alloys cut differently from mild steel and from each other. This page explains why, and how we hold ±0.005 mm on aluminium, stainless, titanium, and nickel parts. Read it to judge which alloy suits your part, and when a design should change instead of the process.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
Why precise CNC alloy processing is not one process
Alloy is a family, not a material. 6061 aluminium, 316L stainless, Ti-6Al-4V, and Inconel behave like four different trades on the same machine. Each has its own cutting speed, its own way of absorbing heat, and its own way of failing.
That is why precise CNC alloy processing is really a set of decisions made before the first cut. Tool grade, feed per tooth, coolant pressure, and workholding all follow from the alloy. Get the order wrong and the part moves, burns, or work-hardens.
The common thread is control. Alloys with high strength-to-weight ratios are also the ones that fight back: they resist cutting, they conduct heat poorly, or they spring back after the tool passes. Control those three behaviors and the rest of the process becomes predictable.
We machine 127 high-precision CNC machines across 3 wholly-owned plants, 16 of them simultaneous 5-axis centers. The alloy still decides the strategy. The machine only executes it.
- 1Soft and gummyAluminium 6061 and 5052 build up on the tool edge if rake angles are wrong.
- 2Hard and abrasive17-4PH and 440C wear tools fast and need lower surface speed.
- 3Heat-trappingTitanium and Inconel push heat into the tool instead of the chip.
- 4Spring-backThin magnesium and aluminium walls deflect under cutting force.
Chip formation: the first thing that changes
A chip tells you whether the cut is working. On 6061-T6 at 3,000–6,000 rpm, a proper chip is bright and curled, and it breaks cleanly. A dull, stringy chip means the feed per tooth is too low and the tool is rubbing.
Stainless 304 and 316L work-harden the moment the tool stops cutting and starts sliding. Feed per tooth below 0.05 mm lets the surface harden under the edge, and the next pass cuts hardened metal instead of soft metal. This is the single most common cause of tool failure on stainless.
Titanium TC4 behaves differently again. Its thermal conductivity is low, so heat stays at the cutting edge. A 12 mm carbide end mill running dry will fail in minutes. We flood the cut at 40–70 bar through the tool, keep surface speed near 40–60 m/min, and accept a shorter tool life rather than a burned part.
Nickel alloys such as Inconel take this further. Surface speed drops to 20–35 m/min, depth of cut stays shallow, and the tool path avoids dwelling in one spot. Every second the tool spends rubbing is a second the work-hardened layer grows.
- 1Aluminium 60612,000–8,000 rpm, feed 0.1–0.3 mm/tooth, high rake, polished flutes.
- 2Stainless 304/316L800–1,500 rpm, feed 0.08–0.15 mm/tooth, never dwell.
- 3Titanium TC4300–600 rpm, feed 0.05–0.12 mm/tooth, high-pressure coolant.
- 4Inconel150–350 rpm, feed 0.05–0.1 mm/tooth, sharp edge, no recutting.
Thermal control and why alloys hold size differently
Aluminium expands about twice as much as steel per degree. A 200 mm aluminium part measured at 30 °C can sit 0.05 mm larger than the same part measured at 20 °C. That is ten times our tolerance band, so we do not measure hot parts.
Stainless 316L expands less but conducts heat poorly, so the cutting zone stays hot and the part stays cool. The tool wears, the part does not grow much. That sounds easier, and it is, until the thin wall gives way under the cutting force.
Titanium sits between the two in expansion but traps heat worse than either. A titanium flange can finish within tolerance and still distort when it cools overnight. We rough, let the part rest, then finish. Sometimes the rest is the most important step in the process.
The practical rule: for tight-tolerance alloy work, decide the measurement temperature before you decide the cutting parameters. Everything else follows from that number.
- 1Measure at 20 °CLet parts stabilize before final inspection whenever the tolerance is under ±0.01 mm.
- 2Rough, rest, finishStandard sequence for titanium and thin-wall stainless.
- 3Coolant on the partFlood cooling stabilizes the workpiece, not just the tool.
- 4Thermal driftLog spindle and ambient temperature through long cycles.
Fixturing alloy parts without marking them
Aluminium and magnesium mark easily. A steel clamp on a finished aluminium face leaves a witness mark that no polishing removes without changing the dimension. Soft jaws, machined to the part profile, solve most of this.
Thin-wall alloy parts need support, not force. We use low-melt fixturing, vacuum chucks on flat plates, and sacrificial tabs on parts under 2 mm wall thickness. The goal is to hold the part without squeezing it out of shape.
For 5-axis work on complex alloy housings, we machine soft jaws in the same setup as the part. That keeps the locating surfaces concentric within ±0.005 mm and removes one re-clamping error.
Titanium and stainless can take more clamping force, but they spring back. A part that measures correctly under clamp pressure may open up 0.02–0.05 mm once released. We verify after unclamping, not during.
- 1Soft jawsMachined to the part profile, replaced when worn.
- 2Vacuum chucksFor thin flat plates and large aluminium panels.
- 3Sacrificial tabsFor walls under 2 mm and small magnesium parts.
- 4Release checkMeasure again after unclamping before final sign-off.
Surface finish and what each alloy can actually reach
Surface finish depends less on the machine and more on the alloy. Aluminium 6061 and 7075 reach Ra 0.2–0.8 μm with a sharp tool and a clean pass. Stainless tends to smear, so Ra 0.8–1.6 μm is a realistic as-machined target.
Titanium will finish well but the tool wears during the pass, so the last 20 mm of a long cut can be rougher than the first. We plan tool changes before the finish pass, not during it.
Magnesium AZ31B and AZ91D cut fast and finish bright, but chips are a fire risk. We keep them dry, use a dedicated dust collection setup, and never mix magnesium chips with other metals.
If a drawing calls for Ra 0.2 μm on 316L, that is usually a polishing operation, not a milling one. We say so before quoting, because quoting a milling pass that cannot reach the number wastes everyone's week.
- 1AluminiumRa 0.2–0.8 μm as machined with sharp carbide.
- 2StainlessRa 0.8–1.6 μm as machined; finer needs polishing.
- 3TitaniumPlan a fresh tool for the finish pass.
- 4MagnesiumBright finish, strict chip handling.
How five common alloys behave in precise CNC alloy processing
Surface speed and feed are starting points for carbide tooling, not fixed rules.
| Alloy | Surface speed | Typical tolerance | Main risk |
|---|---|---|---|
| 6061-T6 aluminium | 300–600 m/min | ±0.005 mm | Built-up edge, thermal growth |
| 304 / 316L stainless | 80–150 m/min | ±0.005 mm | Work hardening, smearing |
| Ti-6Al-4V (TC4) | 40–60 m/min | ±0.005 mm | Heat at the edge, distortion |
| 17-4PH stainless | 60–120 m/min | ±0.005 mm | Tool wear, hard spots |
| Inconel | 20–35 m/min | ±0.01 mm | Notch wear, work hardening |
Which alloy should you actually specify?
If the part is a housing or bracket and the load is moderate, 6061-T6 gives you the tightest tolerance and the best finish at the lowest risk. Choose stainless only when corrosion or strength demands it, and choose titanium or Inconel only when temperature, weight, or chemistry leaves no alternative, because both cost more in time than in material.
Questions engineers ask before releasing an alloy part
Why did my aluminium part measure oversize after machining?
The part was almost certainly measured while still warm from cutting. Aluminium expands roughly twice as much as steel per degree Celsius, so a 200 mm part can read 0.05 mm large at 30 °C.
Let it stabilize to 20 °C in the inspection room, then measure. If the reading is still out, the issue is usually tool deflection or a loose fixture, not thermal growth.
Can you hold ±0.005 mm on titanium?
Yes, on features that are rigid enough and reachable with a short tool. Thin titanium walls below 1.5 mm are a different problem because the material deflects under cutting force.
For those we rough, stress-relieve where the drawing allows, then finish with light passes. We will tell you at quoting stage if a feature cannot hold the tolerance as drawn.
Does the alloy change the lead time?
The machining itself runs at similar throughput for aluminium and stainless. Titanium and Inconel cut three to five times slower because surface speed is much lower, so cycle time grows.
Standard parts ship in 3–5 days once production starts. We confirm the schedule per alloy at quoting.
What surface finish can you reach on 316L stainless?
As machined, Ra 0.8–1.6 μm is realistic and repeatable on 316L. Tighter than that usually means a secondary polishing operation.
If your drawing calls for Ra 0.2 μm on stainless, we will quote the polishing step separately so you can see what the finish actually costs.
Do you cut magnesium alloys?
Yes. We machine AZ31B and AZ91D with strict chip control, dry cutting, and dedicated collection. Magnesium chips are a fire risk if they mix with other metals or sit wet.
If your design allows aluminium instead, that usually removes the handling overhead. If it does not, we machine magnesium on a controlled setup.
How do you inspect alloy parts before shipment?
Every part gets a raw material check, in-process monitoring, and final inspection. Reports are available on request, including dimensional data for critical features.
For tight-tolerance alloy work we record the inspection temperature, because a measurement without that context is only half a data point.
Send an alloy drawing and get a real answer
Upload your part and we will return a quotation with free DFM analysis within 12 hours, including alloy-specific notes on tolerance, finish, and risk.
12-hour quote100% inspectionNo MOQNDA on request
More alloy machining notes
We publish setup notes, tooling trials and inspection data from the factory floor.
