Understanding CNC Citation Formulas: A Beginner's Guide

• 1
  What a CNC citation formula actually models
• 2
  Material cost is a yield problem, not a catalogue price
• 3
  Cycle time: the numbers that drive the machining line
• 4
  Setup, fixturing and how orientation multiplies cost
• 5
  Tolerance and surface finish: where quotes diverge most
• 6
  How quantity and lead time change the same formula
• 7
  Which cost driver moves the quote, and when it stops mattering

A citation formula is a cost model, not a pricing rule. It lists the physical work a part requires, then multiplies that work by rates. For a simple turned bushing, the model is short: bar stock, one setup, a few minutes of cycle time, a quick check. For a five-axis aerospace bracket, the same model grows long because setups, fixturing and inspection each add their own line.

Understanding CNC citation formulas starts with one idea. Price is not a property of the part. It is a property of the process chosen to make the part. Change the stock size, the fixture, the tolerance or the batch size and the number moves, even though the drawing has not changed.

Most shops build a quote from five stacks: material, machining time, setup and programming, finishing, and inspection plus overhead. Each stack has its own drivers. Material follows volume and yield. Machining time follows removal volume and cutting parameters. Setup follows how many orientations the part needs. Finishing and inspection follow the drawing's tolerance and surface callouts.

• 1
  MaterialStock cost divided by parts per bar or plate, with scrap added back.
• 2
  MachiningCutting minutes multiplied by the machine's hourly rate.
• 3
  SetupProgramming, fixturing and first-article time spread over the batch.
• 4
  Finishing and inspectionCoating, deburr and the measurement plan the drawing forces.

The material line in a quote rarely equals the weight of the finished part times the metal price. It equals the stock you must buy, divided by what you can ship. If a part weighs 0.4 kg but is cut from a 1.6 kg block, the buyer pays for 1.6 kg of metal and the shop absorbs the chips. That ratio is the yield.

Yield is decided by geometry. A shallow pocket in a thick plate wastes more than the same pocket in a near-net forging. Wall thickness matters too: a 2 mm wall in aluminium 6061 can be machined from solid, but the same wall in titanium TC4 may distort and need a stress-relief step, which adds cost without adding metal.

Grade choice moves the number in two directions. Aluminium 6061-T6 is cheap and machines fast. Stainless 316L costs more and cuts slower, so the machining line grows as well. Inconel and titanium push both lines up at once. For most parts the grade is fixed by the application, so the real lever is stock form: bar, plate, tube or near-net shape.

Ask for the stock size used in the quote. If the shop quotes a 200 × 200 × 60 mm block for a part that fits in a 200 × 200 × 40 mm plate, the extra 20 mm is pure cost with no function.

• 1
  Near-net stockCuts chips and cycle time when the shape allows it.
• 2
  Oversize plateCommon hidden cost on low-quantity work.
• 3
  Difficult alloysRaise both material and cutting cost.

Cycle time is removed volume divided by material removal rate. That is the whole model. Everything else adjusts it. A 50 cm³ pocket in aluminium 6061 at a healthy removal rate is a few minutes of roughing. The same pocket in 316L stainless can take three to five times longer because the cutting speed drops and the tool must be more conservative.

Feed and speed come from surface speed and chip load. Surface speed depends on the tool material and the work material. Carbide in aluminium runs fast. Carbide in titanium runs slow and floods with coolant. An engineer quoting the job does not guess; the shop's CAM post and tool library already hold tested values for each material pair.

Tool changes and rapids add real minutes. A part with 40 small pockets may cut for only 6 minutes but spend another 4 minutes moving between features. On a five-axis machine, rotary moves and reorientation add more. This is why a geometrically simple part with many features can cost more than a large smooth one.

The most useful question you can ask is where the time goes. A shop that separates roughing, finishing and drilling minutes gives you a lever. If roughing dominates, a different stock form helps. If finishing dominates, the surface callout is the driver, not the geometry.

• 1
  Roughing timeControlled by removal volume and stock form.
• 2
  Finishing timeControlled by surface area and Ra callout.
• 3
  Non-cut timeTool changes, rapids and reorientation.

Every time the part must be re-clamped, the quote gains a setup. A part machined in one orientation from one face needs one setup. A part with features on all six faces needs several, or a five-axis machine that reaches them without re-clamping. The setup count is often the largest single difference between two otherwise similar quotes.

Fixturing is the second half of that line. Soft jaws are cheap. A custom fixture with locating pins and a vacuum plate is not, and its cost must be spread across the batch. On a one-off prototype the fixture can cost more than the machining. On a 10,000-part run the same fixture is nearly free per part.

This is the point where batch size changes the model, not the part. Setup and programming are fixed costs per order. Material and cycle time are variable per part. As quantity rises, the fixed stack flattens against a growing variable stack, and the unit price falls. The curve is steep at first and then almost flat.

A practical check: ask how many setups the quote assumes. If a part with features on three faces is quoted at one setup, either the shop has a five-axis machine that can reach them or the quote is optimistic.

• 1
  One setupCheapest; needs reach from a single orientation.
• 2
  Multi-setupAdds fixturing and re-datum risk.
• 3
  Five-axisTrades machine rate for fewer setups.

Tolerance is not a single number on a drawing. It is a decision about which features are critical. A general block tolerance of ±0.1 mm costs almost nothing extra. A ±0.005 mm bore for a bearing seat costs more, because it needs a separate finishing pass, a temperature-stable measurement, and possibly a rework loop.

The same logic applies to surface finish. As-machined surfaces at Ra 1.6–3.2 μm come straight off a normal finishing pass. A Ra 0.8–1.6 μm callout on a sealing face may need a finer tool and a slower pass. Ra 0.2–0.8 μm usually means lapping, polishing or a secondary operation, and that leaves the machine entirely.

The trap is applying a tight callout across the whole drawing. A part may need one true bore at ±0.005 mm and every other face can sit at ±0.1 mm. Marking the tight callout only where it functions keeps the quote honest and the inspection plan short.

Inspection follows tolerance. A 100% dimensional check with a CMM report on every feature is a different cost from a first-article check plus spot checks. If your drawing does not require the full report, saying so removes work from the quote.

• 1
  Tight but localKeep ±0.005 mm on functional features only.
• 2
  Fine finishBelow Ra 0.8 μm often leaves the CNC machine.
• 3
  Inspection scopeFull CMM report is a real cost line.

Quantity acts on the fixed part of the model. Programming and fixturing do not repeat per part. On a small batch those costs dominate and the unit price looks high. On a long run they vanish into the per-part rate. This is why a quote for one prototype and a quote for 10,000 pieces can differ by an order of magnitude.

The curve is not linear. The first drop from 1 to 10 pieces is large. From 1,000 to 2,000 the change is small, because most of the cost is now material and cycle time. If a supplier offers a big discount at 2,000 pieces when the model says the curve is already flat, the first quote was probably padded.

Lead time enters differently. Standard routing lets a shop batch similar jobs and share setups. A compressed schedule forces a dedicated setup and may force overtime or a second machine. Both add cost. The parts themselves are identical; the schedule is what changed.

For a beginner, the practical rule is simple. Fix the drawing, then vary only quantity or only lead time, never both at once. Otherwise you cannot tell which variable moved the price.

• 1
  Fixed per orderProgramming, fixture, first article.
• 2
  Variable per partMaterial, cycle time, consumables.
• 3
  Rush routingDedicated setup and overtime raises cost.