Skip to content
Millcrown Request a quote

End mill speeds and feeds: the math that starts the conversation

Sooner or later, every buyer fields the question: what speeds and feeds does this cutter run? Here is the honest version of that answer - the standard formulas, one worked example, and the frame that keeps the numbers honest.

The question behind the question

A distributor stocking a new end mill line, a job shop quoting a job, an online seller writing a listing - all of them get asked the same thing by their customers: what speed and what feed? The honest answer has two parts. Part one is the arithmetic, and the arithmetic is public: two short formulas that turn a cutting speed and a chip load into spindle RPM and feed rate. Part two is the caveat, and the caveat is not fine print - it is the whole truth of the matter. Every number the formulas produce is a starting point, adjusted to the machine, the setup and the workpiece. It is never a performance promise. This note covers part one properly, and states part two in a way you can repeat to your own customers with a straight face.

Four terms, kept separate

Most confusion about speeds and feeds is vocabulary confusion. Four terms do all the work, and each one answers a different question.

TermWhat it answersWhere it comes from
Cutting speed (SFM)How fast the edge moves through the material, in surface feet per minuteThe material family sets the starting band; coating and geometry shift it
Spindle speed (RPM)How fast the tool spinsCalculated from SFM and diameter with the standard formula
Chip load per toothHow much material each cutting edge takes per passMaterial, diameter, flute count and rigidity set the starting band
Feed rate (IPM)How fast the tool travels across the workCalculated from chip load, flute count and RPM

The pattern worth noticing: the two quantities a shop chooses are cutting speed and chip load. The two quantities a shop programs are RPM and feed rate. Charts that hand you RPM and feed directly are doing the first two steps for you - useful, but you should know what they decided on your behalf.

The two formulas

Both are industry-standard, in US units, and worth memorizing:

RPM = (SFM x 3.82) / diameter in inches. The 3.82 is a unit-conversion constant, nothing more - it turns surface feet per minute and a diameter in inches into revolutions per minute. It does not depend on the material, the tool or the brand of anything.

Feed rate = chip load x number of flutes x RPM. Each edge takes its chip load once per revolution, so the whole tool takes chip load times flute count per revolution, and RPM converts that into inches per minute.

One worked example, start to finish

Take a 1/2 inch, four-flute carbide end mill, and pick an illustrative cutting speed of 500 SFM for the sake of the arithmetic. The diameter is 0.5 inch:

RPM = (500 x 3.82) / 0.5 = 1,910 / 0.5 = 3,820 RPM.

Now pick an illustrative chip load of 0.002 inch per tooth:

Feed = 0.002 x 4 flutes x 3,820 = 0.008 x 3,820 = 30.6 IPM.

Both starting values were chosen for clean arithmetic. Real starting bands depend on the material family, the coating, and the tool - and this site keeps its speeds and feeds reference deliberately qualitative for that reason. What the example shows is the shape of the decision: one cutting speed, one chip load, two multiplications, and you have program numbers.

Why the first number is rarely the final number

The formulas assume a perfectly rigid, perfectly round, perfectly held tool. Real shops know better. A long tool sticking far out of the holder flexes and needs lighter engagement. A worn collet lets the tool run with runout, which loads one edge unfairly. A thin-walled part rings no matter what the chart said. An old machine with backlash argues about cut direction before it argues about speed. None of this makes the formulas wrong; it makes them a starting point. The shop listens - to the cut, to the chips, to the finish - and adjusts from the calculated baseline. The holder and toolholding setup alone can move the workable numbers more than any chart row.

What a supplier can honestly put in writing

Here is the line the desk holds, and that we suggest buyers hold their own suppliers to. A supplier can confirm in writing: the geometry (diameter, flute count, helix, corner style, length of cut), the substrate, the coating family, the runout acceptance and the packing. That is the spec sheet, and every line of it is checkable. A supplier cannot honestly promise that a given tool will run at a given speed in a given part, because that outcome belongs to the machine, the setup and the workpiece as much as to the tool. When a catalog sentence promises performance before it has seen your part, read that as marketing and keep reading the sheet.

What to ask for instead

Ask the supplier for the geometry lines, confirmed per program. Ask for the material family the tool was ground for - the size series and flute count follow from that. Ask for samples so your shop can run its own starting-point tests on its own machines. That division of labor is the honest one: the supplier owns the geometry, the shop owns the numbers that emerge on its floor. The desk coordinates exactly this exchange - spec sheets and samples from the partner factories, starting-point references for the general cases, and program terms confirmed in writing before anything ships.

Two mistakes the arithmetic prevents

Worth stating what the formulas are actually for, because their value is less about the answer than about the mistakes they make visible. The first is proportion error. A buyer who moves from a 1/2 inch tool to a 1/4 inch tool and keeps the same RPM has halved the cutting speed without noticing - the formula makes the change explicit, and the cut usually announces it too. The second is flute-count error: programming a feed tuned for four flutes onto a two-flute tool doubles the chip load on every edge, which is how healthy programs break healthy tools. Neither mistake survives five seconds of arithmetic. That is the honest case for learning the math even in an era of CAM defaults: the calculation is the check on the defaults, and the person who can run it is the person who notices when a chart, a salesman or a reposted table does not apply.

Where to go deeper

The formula gets its own walkthrough with more worked examples in the RPM formula note, and the second half of the calculation - chip load - gets the same treatment in the chip load guide. Between the two, the arithmetic is covered. The judgment, as always, stays on your shop floor, where it belongs.

Questions buyers ask

Is there one chart that works for every machine?

No. Published charts are starting points built on assumptions about rigidity, holder and coolant that may not match your setup. The reliable method is to calculate a baseline with the standard formulas, run a test cut, and adjust from what the machine tells you.

Does the coating change the numbers?

It can shift the starting band for cutting speed - a TiAlN-family coating, for example, is associated with dry machining of steels - but the formulas themselves never change. Coating is a spec-sheet line, confirmed per program.

What if my spindle cannot reach the calculated RPM?

Small diameters often calculate to more RPM than a machine has. The usual responses are to run at the spindle maximum and accept a lower effective cutting speed, or to use lighter radial engagement strategies. Both are shop decisions, not tool defects.

What should I ask a supplier for, if not speeds?

Geometry, substrate, coating family, runout acceptance and packing, all in writing per program. Those lines are checkable at inspection. Performance outcomes are not checkable at inspection - they belong to the machine and the setup.

Send the geometry, the material you cut and a quantity per specification. The desk replies in writing with a sourcing plan, the reference MOQ and the drawing questions that need deciding.

Request a quote