How to Calculate CNC Machining Time
This guide shows engineers and buyers how to calculate CNC machining time for milling and turning parts, from single features to a full cycle. Read it and you can build a defensible estimate before you send an RFQ, and spot the numbers that do not add up.

In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
Five things to know before you estimate
What CNC machining time includes
CNC machining time is the clock time a part occupies a machine, from raw stock to a finished, inspected component. For a milling job it covers the cut itself, the rapid moves between features, tool changes, and any probing. For a turned part it also covers bar feed and part transfer.
Engineers often quote only the cutting minutes and then wonder why the shop's number is higher. The gap is the non-cut portion, and on small parts it is usually the larger share.
A useful estimate splits the total into four buckets: cut, non-cut motion, tool change, and setup. Each bucket has its own drivers, so each can be argued about separately when a quote looks high.
Add inspection and any post-processing on top if the part needs anodizing, plating, or heat treatment. Those are calendar days, not spindle minutes, and they should be tracked apart from the machining estimate.
- 1Cut timeToolpath length divided by feed rate, summed over every operation.
- 2Non-cut motionRapids, retracts, and safe-Z moves between features.
- 3Tool change and setupPer cycle and per batch respectively; setup does not shrink with volume.
- 4Inspection and finishingCMM checks, deburring, anodizing, plating, and their queue time.
The basic formula for calculate CNC machining time
The starting point is simple: cutting minutes equal total cutting distance in millimeters divided by feed rate in millimeters per minute. If a 2,400 mm contour runs at 600 mm/min, that feature takes 4 minutes.
Feed rate comes from the tooth load and the spindle speed. For a 4-flute carbide end mill at 0.05 mm per tooth and 8,000 rpm, feed is 0.05 × 4 × 8,000 = 1,600 mm/min. Cut that in half for finishing passes and the same path costs twice as much.
For turning, use feed per revolution instead. At 0.2 mm/rev and 1,200 rpm, the tool advances 240 mm/min. A 300 mm long pass then takes 1.25 minutes.
Rapids are not free either. A machine running at 30,000 mm/min covers 500 mm of rapid in one second, but acceleration and deceleration at each corner means the real average is often closer to 15,000–20,000 mm/min.
- 1MillingCut minutes = path length (mm) ÷ feed (mm/min).
- 2TurningFeed (mm/min) = feed per rev (mm/rev) × rpm.
- 3DrillingCut minutes = depth ÷ (feed per rev × rpm).
- 4Pocket clearingUse stepover width to get the real number of passes, not one path.
Setup, fixturing, and non-cut time
Setup is charged per operation, not per part. A vise-mounted 3-axis job with three faces to machine needs three setups. A 5-axis part held once in a self-centering vise needs one. That difference can be 30 to 60 minutes on the first article.
Fixturing is where new designs lose days. Soft jaws, a custom plate, or a vacuum fixture can take 2 to 8 hours to design, cut, and prove out. Repeat orders reuse the fixture, so the per-part setup drops sharply from the second run onward.
In-cycle probing adds 10 to 40 seconds per feature but removes a manual check later. On tight-tolerance work at ±0.005 mm, that trade usually pays for itself.
Batching matters too. Running 50 parts from one setup spreads the setup minutes across 50 pieces. Running 5 parts does not, so small quantities always carry a higher per-part overhead.
- 1Count setups firstEvery new face or datum needs its own setup and its own error stack.
- 2Fixture design is real workBudget 2–8 hours for a custom fixture on a new geometry.
- 3First article costs moreSetup and prove-out dominate the first part, not the cut.
- 4Batch to dilute setupThe same setup spread over more parts lowers unit cost.
How material changes the estimate
Aluminium 6061 cuts fast. Surface speeds of 300 to 500 m/min are normal with carbide, and a 12 mm end mill can run at 1,500 to 3,000 mm/min in roughing without chatter.
Stainless 304 and 316 work-harden, so feeds stay lower and coolant is mandatory. Expect surface speeds around 100 to 150 m/min and roughly 40 to 60% of the aluminium feed rate for the same tool.
Titanium Ti-6Al-4V is worse. Surface speed drops to 40 to 70 m/min, and tool life becomes the limiting factor rather than spindle speed. Many shops change inserts mid-cycle on long titanium parts, which adds minutes that never appear in the formula.
Tool steel and Inconel push surface speeds to 25 to 50 m/min with coated carbide. On these materials, the cut time estimate is only as good as the tool-life assumption behind it.
- 1AluminiumRough at 1,500–3,000 mm/min with a 12 mm carbide tool.
- 2Stainless 304/316Run 40–60% of the aluminium feed; keep the tool moving.
- 3Titanium Ti-6Al-4V40–70 m/min surface speed; watch insert wear.
- 4Inconel and tool steel25–50 m/min; expect tool changes inside the cycle.
Six steps to build the estimate
- 11. Group the featuresList every pocket, hole, contour, and face. Group by tool so you can count tool changes. A part with 40 holes in 4 sizes needs 4 tools, not 40.
- 22. Set the cutting parametersPick surface speed and feed per tooth from the material and tool. For aluminium 6061 at 0.05 mm/tooth, 4 flutes, 8,000 rpm, feed is 1,600 mm/min. Write the number down for each tool.
- 33. Measure the cutting pathPull the path length from CAM for each tool. For a pocket, multiply the number of passes by the pass length. Do not use the part outline; use the toolpath.
- 44. Divide for cut minutesCut minutes = path length ÷ feed. Add a 10–15% factor for acceleration and corner slowdowns. Sum across all tools.
- 55. Add the non-cut minutesRapids at an effective 15,000–20,000 mm/min. Tool changes at 4–8 seconds each. Probing at 10–40 seconds per feature. Add these to the cut total.
- 66. Add setup and divide by quantityAdd 30–60 minutes per setup, plus fixture time on a first article. Divide the setup total by the batch size to get the per-part share. Check the result against chip volume.
Typical parameters and what they do to the estimate
Ranges are starting points for carbide tooling. Confirm with your own tool data before quoting.
| Material | Surface speed | Feed per tooth | Effect on estimate |
|---|---|---|---|
| Aluminium 6061 | 300–500 m/min | 0.05–0.15 mm | Fastest; roughing dominates the cut |
| Stainless 304 | 100–150 m/min | 0.03–0.08 mm | 40–60% of aluminium feed; coolant required |
| Titanium Ti-6Al-4V | 40–70 m/min | 0.02–0.06 mm | Tool life limits the cycle, not rpm |
| Steel 4140 | 80–120 m/min | 0.04–0.10 mm | Moderate; watch work hardening |
| Inconel 718 | 25–50 m/min | 0.02–0.05 mm | Slowest; expect mid-cycle tool changes |
| Brass C36000 | 200–400 m/min | 0.05–0.15 mm | Very fast; good for small runs |
Questions engineers ask next
What is a realistic machining time per part?
It depends on size and feature count more than on material. A simple 50 × 50 × 20 mm aluminium bracket with two setups often runs 3 to 8 minutes of cut plus similar non-cut minutes.
A 200 mm 5-axis housing with 30 features can run 45 minutes to 3 hours. Ask for the split between cut, non-cut, and setup so you can compare like with like.
How do I estimate a new design with no CAM data?
Start from a similar part you have already made and scale by path length and feature count. A part with twice the pocket area and the same tool roughly doubles the cut minutes.
Then add 20 to 30% for the unknown. First articles always carry extra prove-out, and a design that has never been machined will find at least one problem.
Does material type change the estimate more than geometry?
Both matter, but geometry usually wins. A 30-minute aluminium part in 316 stainless can become 60 to 90 minutes. A geometry change that removes half the pocket volume cuts the time in half regardless of material.
When both change at once, estimate them separately. Change the feed for material first, then re-measure the path.
Can setup time be reduced?
Yes, and it is usually the cheapest saving. Design datums so the part can be machined from fewer faces. Move tight tolerances onto the same face. Allow a clamping area that does not need machining.
Running a 5-axis cycle held once can remove two setups entirely compared with a 3-axis route, which often saves 60 to 120 minutes on a first article.
Is 5-axis machining always faster?
No. For a flat plate with holes on one face, a 3-axis machine is faster and cheaper because there is nothing to index. 5-axis pays off when the part has features on many faces, deep pockets with short tools, or complex contours.
The honest test is setup count. If you can go from three setups to one, 5-axis usually wins. If you were already at one setup, it may not.
How accurate does my estimate need to be?
Within 10 to 20% is good enough for budgeting and make-or-buy decisions. Below that, you are spending engineering hours to save minutes.
For a firm quote, the shop will re-run the numbers in CAM anyway. What helps them most is a complete drawing, a material callout, and a clear tolerance stack.
What makes a quote look too high?
Usually one of three things: more setups than you assumed, a tolerance that forces in-process probing, or a surface finish that needs a separate finishing pass at low feed.
Ask which one it is. A shop that can name the driver can also tell you what drawing change would remove it.
Get a machining time you can check
Send your drawings and we will return a quotation with a free DFM analysis within 12 hours, broken down so you can see where the minutes go.
12-hour quote100% inspection±0.005 mm