Why SFM, not RPM, is the material question
Revolutions per minute belong to the spindle; surface feet per minute belong to the workpiece. Cutting speed describes how fast the edge moves through the material, and the material - its hardness, its abrasiveness, how it conducts heat, how it work-hardens - is what decides which speeds it will tolerate and which it will punish. That is why every speeds-and-feeds conversation starts with the material family and works toward RPM through the standard formula, covered with worked examples elsewhere on this blog. A buyer who knows the relative bands can sanity-check any chart a supplier or customer produces; a buyer who only memorizes RPM rows cannot.
The relative bands, honestly stated
Published SFM charts disagree with each other, often by wide margins, because each was built on different assumptions about tooling, coolant and rigidity. What survives across all of them is the ordering - which families run fast, which run slow, and what tends to move a family's band. That ordering is what the table below states, and states qualitatively on purpose.
| Material family | Relative starting band | What the band reflects | Typical tooling notes |
|---|---|---|---|
| Wrought aluminum alloys | High | Soft, conductive, forgiving of speed when chips evacuate well | High flute counts for finishing, polished uncoated flutes against built-up edge; see the aluminum selection |
| Low-carbon steel | Mid | Gummy behavior wants a coated edge and steady load | Coated carbide, climb direction, steady chip load; see the steel selection |
| Alloy and medium-carbon steels | Below low-carbon | More strength and abrasiveness than plain carbon grades | TiAlN-family coatings common, rigid setups preferred |
| Austenitic stainless | Lower | Work-hardening and heat that stays at the edge | Keep the edge engaged and shearing; see the stainless selection |
| Hardened tool steels | Much lower | Hardness itself caps the cutting speed | Rigid everything; geometry choices per the hardened steel note |
| Titanium alloys | Much lower | Poor thermal conductivity keeps heat at the cut | Low bands, high rigidity, steady engagement; see the titanium selection |
| Cast iron | Mid, family-dependent | Graphite helps, abrasive skin and dust punish | Edge preparation matters; dust management often decides dry or air |
A qualitative reference frame, not a numbers table - every real starting point is adjusted to the machine, the setup and the workpiece, and none of it is a performance promise.
Why this page prints no numbers
It would be easy to fill the middle column with figures, and dishonest to do it here. Real cutting-speed recommendations depend on the specific alloy condition, the tool substrate and coating, the coolant plan, the machine's rigidity and the operation - variables this page cannot see. A number pasted out of context reads as authority and behaves as a performance promise, which is exactly the framing the desk refuses. The honest sequence is: take the qualitative band, take the standard formula, calculate a starting point, run it on your machine, and adjust from what the cut tells you. The speeds and feeds reference page sets out that sequence in full.
What moves a band inside a family
The table sorts by family; the finer moves come from the tool and the setup. Coating is the big one: a TiAlN-family coating is associated with running steels dry and hotter, while aluminum usually wants uncoated polished flutes that refuse to let material weld to the edge. Substrate grain matters - fine-grain carbide holds a sharper edge, which shears rather than plows. Geometry matters: helix angle, flute count and corner preparation all change how the edge meets the load. And the setup matters more than any of them - a short, rigid tool in a good holder works at engagement levels a long, slender reach cannot dream of. Each of these lines is specifiable, which is why they belong on the program sheet rather than in general advice.
From band to program, in order
- Name the material family and condition. The table's rows, refined by the actual alloy and hardness you hold.
- Place the qualitative band. High, mid, lower, much lower - the family's honest starting territory.
- Calculate RPM. The standard formula turns the chosen SFM and the diameter into spindle speed, with the arithmetic in the RPM note.
- Choose a chip load and calculate feed. The second standard formula; the reasoning is in the chip load guide.
- Run and adjust. Listen, read chips, change one variable at a time. The machine's verdict outranks any chart.
Reading any chart critically
Charts remain useful objects - the point is to read them like a buyer, not like a believer. Three questions sort a good chart from a decorative one. First: does it state its assumptions? A chart that names the substrate, the coating family and the coolant plan behind its figures can be compared against your conditions; one that publishes bare numbers cannot. Second: is the ordering right? The material rows should descend the way the table above does - aluminum-friendly families at the top, titanium and hardened steels at the bottom - and a chart that breaks that order has either an unusual definition or an error. Third: do the numbers move when the inputs move? A chart that quotes one RPM per material regardless of diameter has skipped the calculation entirely, and every row inherits the omission. None of this rejects charts; it places them where they belong, as someone else's starting points with someone else's assumptions attached.
The same discipline applies to advice travelling the other direction. When a customer asks your business for speeds and feeds, the defensible answer is the one this site gives: the material family's qualitative band, the standard formulas, and the explicit statement that the numbers are starting points to be adjusted on the machine. That answer survives scrutiny because every part of it is true, and it positions your business as the one that taught the customer to think rather than the one that guessed on their behalf.
Test cuts settle what tables cannot
This is where samples earn their cost. A program that will ship thousands of parts deserves a first article run on the machines that will do the work, and the desk builds programs around exactly that rhythm - spec sheet, samples, then bulk. Typical reference ranges for custom grinding samples and bulk production are collected in one place and confirmed in writing per program. If you are matching tools to a set of materials and want the geometry chosen deliberately for each family, send the list through the quote form; the answer will include which family each tool is ground for, and leave the final numbers where they belong - to your test cuts.