Easy Processing: How to Calculate CNC Production Time
This guide explains what actually drives machine hours on a CNC job. It is written for engineers and buyers who need to sanity-check a quoted lead time instead of accepting it. After reading it you can break a part into operations and estimate where the hours go.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
What production time is made of
Machine time is not one number. It splits into setup, non-cutting motion and actual metal removal. Setup covers workholding, zeroing, tool loading and first-article checks. Non-cutting motion covers rapids, tool changes and indexing. Metal removal is the only part that scales with the size of the cut.
When people try to calculate CNC production time, they usually multiply cycle time by quantity and stop there. That works for a run of 5,000 identical brackets. It falls apart on a one-off housing with three setups, because setup can be longer than the cut.
The ratio between setup and cutting is the first thing to check. If setup is 40 minutes and the cycle is 6 minutes, a run of 10 parts spends 460 minutes on the machine, and most of it is not making chips.
A useful habit is to list every operation in order before estimating anything. Face, rough, semi-finish, finish, drill, tap, deburr. Each one has its own tool, feed and time.
- 1Setup timeFixturing, zeroing, tool offsets, first-article inspection
- 2Non-cutting timeRapids, tool changes, rotary indexing, coolant waits
- 3Cutting timeVolume of material removed divided by removal rate
- 4Post-processingDeburring, finishing, cleaning, inspection outside the machine
The formula behind calculate CNC production time
The working formula is production time = setup time + (part volume removed ÷ material removal rate) + non-cutting time + post-processing time. Volume divided by removal rate gives cutting minutes. Everything else is added on top.
Material removal rate for milling is axial depth × radial width × feed rate. For a 12 mm end mill in 6061 aluminium, a typical roughing pass might run 3 mm axial depth, 6 mm radial width and 2,000 mm/min feed. That is 36,000 mm³/min, or 36 cm³/min.
Turning uses depth of cut × feed per revolution × surface speed, expressed as volume per minute. The same logic applies. The numbers change, the structure does not.
Once you have a removal rate, divide the stock volume to be removed by it. A part that starts as a 150 × 100 × 40 mm block and ends at 60 percent of that volume has roughly 240 cm³ to remove. At 36 cm³/min that is about 6.7 minutes of roughing, before finishing passes.
- 1Cutting timeRemoved volume ÷ removal rate
- 2Removal rateAxial depth × radial width × feed
- 3Realistic derateApply 60–75 percent to the ideal number
Why setup time dominates small batches
Every new orientation is a new setup. A part machined on three faces needs at least three fixture positions unless a 5-axis machine reaches them in one. Each position costs the time to clamp, indicate and touch off tools.
On a 3-axis machine, a 30-minute setup per face is normal for a mid-size prismatic part. Three faces means 90 minutes before the first chip. A simultaneous 5-axis centre with a Ø400 mm rotary table can often reach the same faces in one setup, which is why it wins on complex geometry even at a higher hourly rate.
The break-even is quantity-driven. Below roughly 20 parts, setup usually outweighs the cutting saving of a cheaper machine. Above a few hundred parts, cycle time decides everything and the cheapest capable machine wins.
Quick-change zero-point systems and soft jaws cut setup time without changing the cut. They do not reduce cycle time at all. Keep the two separate when you build an estimate.
- 1Count the facesEach new orientation is a new setup
- 23-axisCheap hour rate, more setups
- 35-axisOne setup, higher hour rate
- 4FixturesZero-point plates cut setup, not cycle
Material and hardness shift the numbers
The same part in 6061-T6 and 17-4PH stainless does not take the same time. Aluminium cuts at high surface speed with deep passes. Stainless work-hardens, so light passes and lower feed are safer. Titanium Ti-6Al-4V is worse again because heat stays in the cut and tool life drops.
A practical rule is that removal rate falls by roughly half when moving from aluminium to mild steel, and by another third or more going to stainless or titanium. These are shop rules of thumb, not handbook constants, but they get you within the right order of magnitude.
Hardness matters less than thermal conductivity for tool life. That is why 304 stainless and Ti-6Al-4V both punish the tool even though their hardness numbers look modest.
Inconel and magnesium sit at opposite ends. Inconel is slow, magnesium is fast but needs care with chip handling. Neither is a material where a generic estimate survives contact with the machine.
- 1AluminiumDeep passes, high surface speed
- 2StainlessLight passes to avoid work hardening
- 3TitaniumLow speed, heat stays in the cut
- 4PlasticsFast, but watch chip evacuation and melting
Geometry, tolerances and where time hides
Feature count drives tool changes. Twenty holes with five different diameters means five tools, five offsets and five approach moves. Twenty holes of one diameter is a single canned cycle. The geometry looks similar on a drawing and takes very different time.
Tolerance changes the strategy. A ±0.005 mm bore is not roughed and finished with the same pass. It needs a semi-finish, a measuring cycle and often a spring pass. That may add 30 to 60 percent to the time for that feature alone.
Thin walls are a hidden cost. Below about 1 mm wall thickness in aluminium, deflection forces lighter passes and more spring passes, so removal rate drops even though the tool is unchanged.
Surface finish has the same effect. Going from Ra 3.2 μm as-machined to Ra 0.8 μm adds a finishing pass at reduced stepover. Going to Ra 0.2–0.8 μm may add a separate finishing operation outside the machine.
- 1Tool countEach new diameter is a new tool and offset
- 2Tight toleranceAdds semi-finish and measuring cycles
- 3Thin wallsLighter passes, lower removal rate
- 4Fine finishSmaller stepover, more passes
How machine and tooling change the result
Spindle power and rigidity set the ceiling on removal rate. A small spindle cannot take a deep pass in steel no matter what the feed table says. During any attempt to calculate CNC production time, the machine envelope and spindle rating are hard limits, not soft preferences.
Toolpath strategy matters as much as the tool. Trochoidal and high-efficiency milling keep radial engagement low and axial depth high. That spreads heat and load along more of the flute, which raises removal rate in hard materials.
Tool holding is often ignored. A long reach tool in a standard holder deflects, so the programmer backs off the feed. A shrink-fit holder with the shortest reach lets the same tool run at the planned rate.
Multi-pallet machines and lights-out running change the arithmetic. If a machine runs unattended overnight, a 9-hour cycle does not consume 9 hours of labour, but it still consumes 9 hours of machine time.
- 1Spindle powerCaps depth of cut, especially in steel
- 2ToolpathTrochoidal paths raise removal rate in hard stock
- 3Tool holdingShort reach and shrink fit allow full feed
- 4Unattended runningMachine hours continue without labour
Removal rate and time by material and operation
Typical shop-floor ranges for the same part envelope. Use them to bracket an estimate, not to price a job.
| Material | Roughing rate | Relative time | Watch out for |
|---|---|---|---|
| 6061-T6 aluminium | 30–40 cm³/min | 1.0× baseline | Chip welding at low coolant flow |
| 1018 / 1045 steel | 12–18 cm³/min | 1.8–2.2× | Work hardening on light passes |
| 304 / 316 stainless | 8–14 cm³/min | 2.5–3.5× | Heat and tool wear |
| 17-4PH stainless | 6–10 cm³/min | 3–4× | Post-heat-treat machining only |
| Ti-6Al-4V | 4–8 cm³/min | 4–6× | Tool life, coolant pressure |
| Inconel | 2–4 cm³/min | 6–8× | Very low speed, rigid setup |
| POM / PEEK | 25–35 cm³/min | 0.9–1.2× | Thermal growth on long parts |
Which estimate you should trust
For a one-off or prototype, estimate from setup count and tool count, and treat cutting time as secondary. For a repeat production run, estimate from removed volume divided by removal rate, then derate to 60–75 percent for real tool changes and rapids. If the two methods disagree by more than a factor of two, the part is probably not ready for a firm quote.
Questions engineers ask
Can I calculate CNC production time from the CAM file alone?
The CAM simulation gives you toolpath length and cycle time for the cut. It excludes fixturing, first-article inspection, tool changes that happen outside the program and any rework.
For simple parts the simulation is close. For parts with tight tolerances or thin walls, add 30 to 60 percent for the steps a programmer inserts after the first article.
Why does the quoted time not match my own estimate?
The usual causes are setup count and inspection. A quote may include in-process monitoring on a tight feature, or a separate fixture that your estimate treated as one setup.
Material condition matters too. Pre-hardened or heat-treated stock cuts differently from annealed stock, and the CAM file does not know which one arrived.
Does quantity change the formula?
The cutting-time term scales with quantity. The setup term divides across the batch. That is why a 5-part order and a 5,000-part order can have very different minutes-per-part numbers.
The setup term never disappears entirely. Even a long run needs periodic tool changes and inspection, which is why the derate factor stays in the estimate.
How much time do tight tolerances really add?
For a bore held at ±0.005 mm, expect a semi-finish pass, a measuring cycle and often a spring pass. That is commonly an extra 30 to 60 percent on that feature.
The cost is not only cutting. Gauging and temperature stabilisation before measurement can add floor time that never shows up in the cycle.
Is a 5-axis machine always faster?
No. It is faster when the part needs several orientations, because it removes setups. On a simple part that fits one 3-axis setup, the 5-axis machine often runs at the same rate but at a higher hourly cost.
The advantage appears when setup count drops, not when the cut itself changes.
How accurate can a shop-floor estimate be?
On repeat work with known material and fixtures, a good estimator lands within 10 to 15 percent. On a first-time part with new geometry, 25 percent is realistic.
Treat any estimate tighter than that as a target, not a measurement. The first article is what confirms it.
Send the drawing, get the machine hours back
We review the geometry, count the setups and return a quotation with a DFM analysis within 12 hours. No minimum order quantity, from one prototype to 10,000+ part runs.
12-hour quoteFree DFM analysis100% inspection
More process notes from the shop floor
We publish setup notes, tooling trials and inspection data from the factory floor.
