Choosing a CNC Milling Service for Precise Seating
A seat, pocket, counterbore or locating shoulder is judged by how well two parts mate, not by how the drawing looks. This guide is for engineers and buyers comparing shops, and it lists the numbers that decide whether a seat actually holds: tolerance, flatness, surface finish, material, setup and inspection.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
What decides a seat
Which process suits which seat
Pick the row that matches your seat geometry and yearly volume.
| Seat type | Best process | Typical tolerance | Watch out for |
|---|---|---|---|
| Shallow pocket, open side | 3-axis milling | ±0.02 mm | Floor flatness after tool wear |
| Deep pocket, 2 walls | 3-axis + long reach tool | ±0.02 mm | Tool deflection at 4×D depth |
| Bore and seat on one axis | 4-axis milling | ±0.01 mm | Indexing error between faces |
| Angled or contoured seat | 5-axis milling | ±0.005 mm | Post-processor rounding errors |
| Turned seat with milled flats | Mill-turn center | ±0.01 mm | B-axis clamp repeatability |
| Thin-wall seat, 0.8 mm wall | 5-axis, light radial cuts | ±0.01 mm | Chatter and heat distortion |
| Large frame seat, 2,000 mm | Gantry 3-axis | ±0.05 mm | Thermal drift over long cycles |
| Prototype, 1–20 pcs | 3-axis or 5-axis | ±0.01 mm | No fixture, so datums shift |
The takeaway
Pick the shop that tells you how many setups the seat needs and what it will measure. If the quote does not mention datums or inspection, the price is not comparable.
What a precise seat actually is
A seat is any feature that positions another part: a pocket that holds a bearing, a counterbore that centers a pin, a shoulder that stops a shaft, a recessed ledge that keeps a cover flush. The drawing usually gives a width or diameter, a depth and a surface callout. What it rarely gives is the relationship that matters most, which is how the seat floor sits relative to the bore or face it must align with.
That relationship is why a CNC milling service for precise seating is judged on geometry, not just on size. A pocket can measure perfectly at the top and still rock a bearing if the floor is 0.03 mm out of flat. A shoulder can hit its depth and still let a shaft tilt if the shoulder face is not square to the bore axis.
So when you compare shops, ask how they hold the datum chain. A shop that machines the bore and the seat in one setup removes one source of error before the first chip is cut. A shop that machines them in two setups has to prove the refixturing, and that proof is a report, not a promise.
- 1Function firstDecide whether the seat locates, seals, or only clears. That choice sets the tolerance.
- 2Datum chainList which features must stay related, then ask how many setups they need.
- 3Wear surfaceAny seat that slides or presses needs a finish callout, not just a size.
Tolerance and flatness calls that hold up
Most seat drawings over-tighten the wrong dimension. The width or diameter that presses against a mating part needs the tight call, often ±0.005 mm to ±0.01 mm on small parts. The depth that only sets clearance can usually sit at ±0.05 mm without affecting function. Splitting the call this way lowers cost and removes inspection arguments later.
Flatness is the quiet one. On a 40 mm × 40 mm seat floor, a 0.02 mm bow reduces contact area more than a 0.01 mm size error does. If the seat carries load or transfers heat, put a flatness call on the floor. A general note like 0.02 mm over the seat area is enough for most work. Where a gasket or O-ring seals against the floor, tighten it and add a finish callout.
Squareness between the seat floor and any bore matters for anything that rotates. A 0.01 mm per 25 mm squareness call is realistic on a 5-axis machine with a probe. Asking for 0.005 mm per 100 mm on a part with a 0.8 mm wall is a different problem, because the part moves when you cut it.
- 1Split the callTight on the mating width, looser on depth and clearance.
- 2Flatness over sizePut a flatness note on any load-bearing floor.
- 3Square to the bore0.01 mm per 25 mm is a fair target for rotating parts.
Surface finish and how it changes the fit
Finish changes the effective size of a seat. A milled wall at Ra 1.6–3.2 μm carries visible tool marks, and those peaks sit above the nominal surface. Press a pin into that bore and you are pressing against the peaks. The bore feels tight, then loosens as the peaks wear down. If the joint has to hold torque or preload, that loss of interference is a real failure mode.
For most pressed or bolted seats, Ra 0.8–1.6 μm is enough. It comes from a sharp finishing cutter, correct feed per tooth and a stable setup. Going to Ra 0.2–0.8 μm means a smaller stepover, a longer cycle and often a second operation. That cost is justified when a seal runs against the surface, when a bearing outer ring sits directly in the bore, or when the seat sees sliding contact.
The mistake to avoid is calling a fine finish everywhere. A single Ra 0.4 μm note on a part with eight pockets can double the cycle time for no functional gain. Mark the two or three surfaces that touch, and leave the rest as machined.
- 1As machinedRa 1.6–3.2 μm. Fine for clearance and non-contact faces.
- 2Standard fitRa 0.8–1.6 μm. Covers most pressed and bolted seats.
- 3Seal or bearingRa 0.2–0.8 μm. Budget for extra cycle time.
Material choice and how it behaves in a seat
Aluminum 6061-T6 is the default for seat work that does not see heavy wear. It machines fast, holds ±0.005 mm on small features and takes anodizing without losing size if the shop masks the seat. If the seat slides against steel, hardcoat anodizing adds wear resistance, but it builds 0.02–0.05 mm per surface, so the shop must machine undersize to compensate. Say so on the PO.
Stainless 304 and 316L resist corrosion but work-harden. A seat cut with a dull cutter will glaze the wall, and the next pass rubs instead of cutting. Shops that run stainless seats keep separate tooling and lower surface speed. For a wear surface, 17-4PH in the H900 condition gives higher hardness and still mills cleanly before aging.
Titanium TC4 (Ti-6Al-4V) suits lightweight seats in aerospace and medical work, but heat stays in the cut. The usual answer is high-pressure coolant and conservative radial engagement. Plastics like POM and PEEK move after machining, so a seat held to ±0.005 mm in PEEK is only realistic if the shop stress-relieves the blank and inspects after a settling period.
- 1Aluminum 6061-T6Fast, stable, good for anodized seats. Mask the fit surfaces.
- 2304 / 316LCorrosion resistant. Watch work hardening on the finishing pass.
- 317-4PH H900Harder wear surface. Machine before aging, account for shrink.
- 4PEEK and POMMove after cutting. Stress relieve and inspect later.
How to compare quotes and lead times
Quotes for the same seat often differ by 40 percent, and the difference is rarely the hourly rate. It is the setup count, the inspection plan and whether the shop quoted the finish you asked for. Ask for the number of setups in the quote. A seat with a bore and a shoulder quoted in one 5-axis setup is a different job from the same part quoted in three operations.
Lead time has two halves: the time to first chip and the time to ship. A shop that returns a quotation and a DFM analysis within 12 hours and can start production within 24 hours is set up for short runs. Parts that ship in 3–5 days are normal for this kind of work when material is in stock. If your alloy is not on the shelf, add the mill lead time before you promise a date to your own customer.
Volume should not change the process. With no minimum order quantity, a single prototype and a 10,000-part run can go through the same shop. What changes is the fixture. A prototype seat is often cut from a vise, while a production seat gets a dedicated soft-jaw or vacuum fixture. Ask which one the quote assumes.
- 1Setup countAsk for it in writing. It explains most of the price gap.
- 2Inspection planCMM report or hand tools? The answer changes the price.
- 3Fixture assumptionVise for one part, dedicated fixture for a run.
Step by step: qualifying a shop for seat work
Work through these in order. Each step has a check you can verify without visiting the floor.
- 1Send the drawing with the datum chain markedHighlight the seat width or diameter, the floor flatness and the squareness to any bore. Note which surfaces touch and which only clear. A shop that quotes without this will guess, and the guess shows up in the setup count.
- 2Ask how many setups the seat needsOne setup for related features is the goal. If the answer is three operations, ask what tolerance each refixturing adds. A reasonable answer is 0.01–0.02 mm per extra setup on a well-fixtured part.
- 3Confirm the tolerance splitState the tight call on the mating feature and a looser call on depth. If the shop pushes back with a reason, listen. If they accept ±0.005 mm on every dimension without comment, they may not have read the part.
- 4Match finish to functionAsk for Ra 0.8–1.6 μm on contact surfaces and as machined elsewhere. Request the measurement method: a profilometer trace on the seat floor, not a visual check.
- 5Check material stock and machinabilityFor 6061-T6, 304, 17-4PH or TC4, confirm the shop has run the alloy recently. Ask about coolant pressure for titanium and stress relief for PEEK. Work hardening in 304 is the most common cause of a failed finishing pass.
- 6Require an inspection report on the seatAsk for CMM position of the seat center, flatness of the floor and a diameter trace at two heights. This catches taper in a deep pocket that a single diameter reading hides.
- 7Verify the quality system, not just the certificateISO 9001:2015 and IATF 16949:2016 cover process control. If the seat goes into a medical device, ISO 13485:2016 applies. Ask who signs the final inspection record and what happens when a seat is out of tolerance.
- 8Agree on the first-article processFor a production seat, ask for a first article with the full dimensional report before the run continues. For a prototype, ask for a DFM note within 12 hours so you can fix the drawing before chips fly.
Questions buyers ask about seat milling
Can a 3-axis mill hold ±0.005 mm on a seat?
Yes, on a single feature with a rigid setup and a sharp cutter. The limit is not the axis count, it is the datum chain. If the seat and the bore it must align with are on different faces, a 3-axis machine needs a second setup, and that refixturing typically adds 0.01–0.02 mm of position error.
If the seat is shallow and open on one side, 3-axis milling is the economical choice. If the seat must stay related to a bore on another face, one 5-axis setup is usually cheaper than proving a two-setup stack.
How deep can a milled seat go before tool deflection matters?
Deflection grows quickly past 4× the cutter diameter. A Ø6 mm end mill cutting a 24 mm deep pocket is at the limit. Beyond that, use a necked cutter, a smaller stepdown and a finishing pass with light radial engagement.
Deep narrow seats also trap chips. If the shop does not use through-coolant or peck cycles, the cutter recuts chips and the wall finish drops. That shows up as Ra 3.2 μm or worse on a surface you called at Ra 1.6 μm.
Does anodizing change the seat size?
Yes. Hardcoat anodizing builds roughly 0.02–0.05 mm per surface, and it grows both inward and outward. A seat that measures Ø20.000 mm before coating will not measure Ø20.000 mm after.
Tell the shop which surfaces must stay at size. The usual fix is to machine the seat undersize by the coating thickness and mask the fit surfaces so they stay bare. Clear anodizing builds less but still enough to matter on a ±0.005 mm call.
What should the inspection report actually contain?
For a seat, three things: the position of the seat center relative to the datum, the flatness of the floor, and a diameter trace at the top and bottom of the feature. The two-height trace catches taper that a single reading misses.
Ask for the report as a PDF with the measured values, not a pass or fail stamp. 100% inspection before shipment is the baseline. Raw material check and in-process monitoring should also be part of the record.
How do I handle a seat that must be leak-tight?
Treat the sealing face as a separate feature with its own callouts. Call flatness on the floor, set the finish at Ra 0.8 μm or finer, and specify no visible tool exit marks across the seal path.
If the seat holds a gasket, the floor flatness matters more than the wall diameter. A 0.02 mm bow across a seal path will leak no matter how tight the side walls are. Ask for a leak test before shipment on any seat that carries pressure.
Can one shop handle both the prototype and the production run?
It can, and that usually saves time because the datums stay the same. With no minimum order quantity, a seat can start as one part and scale to a 10,000-part run without a supplier change.
What changes is the fixture. Expect a vise or soft jaws for the prototype and a dedicated fixture for the run. Ask the quote which one it assumes, because the fixture cost is often hidden in the piece price.
Send the seat drawing, get a real answer
Quotation and free DFM analysis within 12 hours. Production can start within 24 hours, and parts ship in 3–5 days.
12-hour quote100% inspectionNo minimum order quantityNDA on request
More from the shop floor
We publish setup notes, tooling trials and inspection data from the factory floor.
