How to Maximize CNC Processing Profits
Profit in a machine shop is not a sales problem. It is a setup, tooling, and scheduling problem. This page explains the five levers that move unit cost, and when each one stops working.
Where machining profit actually leaks
A CNC quote has three cost pools: spindle time, non-cutting time, and the cost of a bad part. Most shops watch the first one. The second and third are where margin disappears. If a 4,000 mm part spends 90 minutes in the fixture and 25 minutes under the cutter, the fixture is the expensive operation, not the milling.
Cutting time scales with material removal rate, and removal rate scales with the rigidity of the whole loop: spindle, tool holder, workpiece, fixture, table. A 12 mm carbide end mill in 6061 aluminium can run at a 6 mm axial depth and a 3 mm radial step. The same cutter in 17-4PH stainless drops to roughly 1.5 mm axial and 1 mm radial before the tool deflects.
So the first question on any job is not which machine is free. It is which machine can hold the part rigidly enough to let the tool run at its intended chip load. Answer that wrong and you pay for it in every pass.
Non-cutting time is easier to control than most programmers admit. Tool changes, rapid moves, and air cutting are all visible in the G-code before the first chip is made.
- 1Spindle timeSet by material, tool, and rigidity. Hard to cut after the fact.
- 2Non-cutting timeSet by programming and fixture design. Easy to cut before the first chip.
- 3Bad-part costScrap, rework, and a second setup. The most expensive pool.
Setup reduction is the fastest lever
Setup is the only cost that is almost entirely within the shop's control. A three-axis machine with a twin vise and pre-set tool offsets can go from last part to first good part in under 20 minutes. The same job on a bare table with paper shims takes an hour and a half.
The mechanism is simple. Every minute spent zeroing a workpiece, touching off tools, and proving a program is a minute the spindle is not making money. On a batch of 50 parts the difference is small. On a batch of 5, it decides whether the job is profitable at all.
Dedicated soft jaws or a modular fixture plate pay back after two or three runs. They also remove the operator's judgement from the process, which is where variability enters.
For five-sided work, a Ø400 mm rotary table with a pre-set zero point turns three setups into one. The part is dialed in once and the table does the repositioning.
- 1Pre-set tool offsetsTouch off on a presetter, not in the spindle.
- 2Dedicated soft jawsPay back after two to three runs on a repeating job.
- 3One setup instead of threeRotary table or trunnion keeps the datum repeatable.
Tool life and tolerance trade-offs
Running a tool harder to save cycle time is a real lever, but it has a ceiling. Past a certain surface speed the cutting edge breaks down through diffusion and notching, and the cost shifts from spindle minutes to insert changes and scrapped parts.
In aluminium, the limit is usually chip evacuation, not tool wear. A high-helix cutter with through-spindle coolant can run at 300 m/min and up. In 316L stainless, the same cutter at the same surface speed will work-harden the surface and fail within minutes. Drop to 120–150 m/min and increase feed per tooth instead.
Tolerance and surface finish set a second ceiling. A Ra 0.8–1.6 μm finish usually needs a separate finishing pass at a smaller radial step. Trying to hit it in one roughing pass means chatter, and chatter means a part that measures in tolerance but fails on assembly.
The useful rule: rough with the biggest tool the geometry allows, finish with the smallest step the tolerance allows. Two passes, two purposes.
- 1Chip evacuation firstRecutting chips damages edges faster than heat does.
- 2Match speed to alloyAluminium and 316L do not share cutting parameters.
- 3Separate rough and finishOne pass cannot do both jobs without chatter.
Batch size, lead time, and machine choice
Lead time is a profit lever because it sets how much work in progress sits on the floor. Parts that ship in 3–5 days free up floor space and cash faster than parts that sit for three weeks waiting for a finishing step.
Batch size interacts with setup cost. If setup takes 30 minutes and cycle time is 8 minutes, moving from 10 parts to 20 parts cuts unit cost by roughly 20%. Moving from 200 to 210 parts changes almost nothing. There is a knee in that curve, and it is worth finding per job.
Machine choice matters on long parts. A 4,000 × 400 × 150 mm travel machine handles a large weldment in one setup, which beats splitting it across two smaller machines and adding a re-fixturing step. On small prismatic parts, a 500 × 500 × 450 mm machine is faster to load and cheaper per hour.
The wrong choice is expensive: a small part on a large machine runs at large-machine rates for no benefit.
- 1Find the kneeDoubling batch size helps at 10 parts, not at 200.
- 2Match travel to part4,000 mm travel for large weldments only.
- 3Shorten finishing queuesAnodizing and plating delays eat lead time.
Cost of poor quality and inspection placement
A scrapped part costs the material, the machine time already spent, the setup time, and the slot in the schedule that is now empty. On a part with 4 hours of cutting, that is a large number. Catching the error at the first operation costs a fraction of catching it at the last.
In-process monitoring is cheaper than final inspection. Checking a critical bore after op 1, while the part is still in the fixture, means the operator can offset and re-cut. Checking it after op 3 means the part cannot be recovered.
For tight work, a 100% inspection before shipment is a cost, not a benefit, unless the process is not capable. Capable processes need sampling. Incapable processes need fixing, not more inspection.
The 99.99% qualification rate on repeating production work comes from process control, not from inspecting every feature twice.
- 1Inspect earlyCatch a bore error while the part is still fixtured.
- 2Fix the processMore inspection does not make an incapable process capable.
Which lever to pull first
Pick the row that matches the job, not the shop.
| Job situation | First lever | What it changes |
|---|---|---|
| 5 parts, 3 setups each | Fixture and 5-axis | Collapses 3 setups into 1 |
| 200 parts, simple geometry | Cycle time and tooling | Setup is already amortized |
| Long weldment over 1,500 mm | Machine travel | Avoids re-fixturing step |
| Tight bore, ±0.005 mm | In-process check after op 1 | Recovers parts instead of scrapping |
| Ra 0.2–0.8 μm finish | Separate finishing pass | Prevents chatter and rework |
| 18-minute cycle, 20-minute setup | Batch size | Find the knee, then stop |
The trade-off in one line
If setup dominates the quote, cut setup time first. If cycle time dominates, cut tool load and batch size. Do not spend money on a faster spindle to fix a fixture problem.
Frequently asked questions
Does a higher spindle speed always reduce unit cost?
No. Spindle speed only helps when the tool, holder, and fixture can absorb the load. Past that point the cutting edge breaks down faster and the savings move to tool replacement.
On aluminium, speed is usually free. On 316L stainless and Inconel, the limit is heat at the edge, not spindle rpm.
When is 5-axis cheaper than 3-axis?
When the part needs four or five faces machined and would otherwise need three or more separate setups.
If the geometry is prismatic and fits one orientation, a three-axis machine with a good fixture is faster to load and cheaper per hour.
How much does batch size affect unit cost?
It depends on the ratio of setup time to cycle time. If setup is 30 minutes and cycle is 8 minutes, doubling from 10 to 20 parts cuts unit cost by roughly 20%.
Once setup is amortized to a few percent of the total, larger batches stop helping and only add inventory.
Is 100% inspection worth the cost?
Only when the process is not capable of holding the tolerance on its own, or when the customer requires it.
A capable process with in-process checks at the first operation usually catches errors earlier and cheaper than final inspection.
What tolerance can be held without a second finishing pass?
As-machined finish runs around Ra 1.6–3.2 μm. Tolerances down to ±0.005 mm can be held at that finish if the setup is rigid.
Finishes of Ra 0.8–1.6 μm and below generally need a separate finishing pass with a smaller radial step.
Does shortening lead time always improve profit?
It improves cash flow and floor space, which matters on repeat work. It does not help if the shortcut adds a rework step.
Shipping in 3–5 days is useful when finishing and inspection are scheduled in parallel, not queued behind production.
Send us the drawing and the batch size
We will return a quotation and a free DFM analysis within 12 hours, with the setup and tooling assumptions written out.
12-hour quote100% inspectionNo minimum order quantity
More from the shop floor
We publish setup notes, tooling trials and inspection data from the factory floor.
