Methodology
How MillCaliber calculates its starting estimates
Every equation, constant, rounding rule and limit the calculator uses, with the reason it was chosen. Nothing here is hidden in the code.
1. Core equations
These are the standard milling relationships (see Harvey Tool's general machining guidelines). D is the cutter diameter in inches and z is the number of flutes.
Spindle speed RPM = 12 × SFM ÷ (π × D) (≈ 3.82 × SFM ÷ D) Surface speed SFM = RPM × π × D ÷ 12 Cutting feed IPM = RPM × z × IPT Removal rate MRR = WOC × DOC × IPM (in³/min)All internal calculation is done in inches. Metric results are converted at display time with 1 in = 25.4 mm, 1 ft = 0.3048 m and 1 in³ = 16.387064 cm³, so imperial and metric show the same underlying estimate.
2. Material baselines
Each material has a target surface speed and a baseline chip load (feed per tooth, meaning the maximum chip thickness) for a 1/4" solid-carbide cutter on a rigid machine. The engagement factor is used only for the engagement guideline in section 7.
| Material | Target SFM | Base chip load, 1/4" (in/tooth) | Engagement factor | Published comparison for 1/4" (in/tooth) |
|---|---|---|---|---|
| 6061-T6 aluminum | 1,200 (unvalidated) | 0.0023 | 1.00 | Chip load: CNC Router Store chart 0.003–0.005 (with lubricant); Productivity RedLine 6061 chart 0.0015 (slot) to 0.0024 (side roughing); UPenn guide 0.002. Surface speed: 600 (Amana slot), 600–1,000 (Productivity, by operation), 600+ (UPenn), 720–1,080 (Helical), about 3,900 at 60,000 RPM (RobbJack, rigid setup). |
| Softwood / MDF | 1,800 | 0.0060 | 0.40 | CNC Router Store chart: softwood 0.009–0.011, MDF 0.011–0.013. |
| Hardwood (oak/walnut) | 1,400 | 0.0045 | 0.60 | CNC Router Store chart: 0.008–0.010. |
| Cast acrylic | 900 | 0.0040 | 0.50 | CNC Router Store chart (hard plastics/acrylic): 0.006–0.009. |
| Free-machining brass (C360) | 600 (unvalidated) | 0.0015 | 1.15 | UPenn guide (brass, not alloy-specific): 300 SFM, chip load 0.002 for a 1/4" tool. No tool-maker chart for C360 has been checked yet. |
How to read these numbers
- Most baseline chip loads are below the cited router chart. For aluminum, softwood/MDF, hardwood and acrylic, MillCaliber's baselines are roughly 55–80% of the low end of the CNC Router Store ranges. Being below a cited example does not make a value safer or more suitable: a chip that is too thin can rub and overheat the cutting edge (see Diagnosing problems). Suitability for any particular tool and machine remains unvalidated. Check your tool maker's data and make a test cut.
- The aluminum target of 1,200 SFM is unvalidated. Published surface speeds for aluminum vary widely with the tool and machine: 600 SFM in Amana's aluminum-cutting slotting chart, 600–1,000 SFM in Productivity's RedLine 6061 chart (600 slotting, 800–1,000 side roughing, 650 finishing), "600+" in the UPenn shop guide, 720–1,080 SFM in Helical's aluminum parameters, and about 3,900 SFM for a 1/4" tool at 60,000 RPM in RobbJack's 6061 data for a rigid setup. 1,200 SFM sits within that range, but no source checked specifies it for the router-class cutter on a desktop machine that MillCaliber assumes. In practice, a 24,000 RPM spindle caps a 1/8" cutter at about 785 SFM, so the target mainly affects cutters of about 1/4" and larger. For a 1/4" cutter, 1,200 SFM gives 18,000 RPM, compared with 15,000 RPM at 1,000 SFM and 9,000 RPM at 600 SFM. Those RPM figures only show how sensitive the result is to the target; they are not proposed replacements. No replacement value has been chosen. Any change will be reviewed separately against tool-maker data for the cutter type MillCaliber assumes, and will not be chosen simply because it is lower or higher. Until then, use your tool maker's surface speed and enter a max RPM that matches it.
- The brass target of 600 SFM is also above the one published example checked (300 SFM, UPenn guide), and is labeled unvalidated. On desktop machines the 10,000 RPM minimum usually sets the spindle speed for brass anyway.
- Engagement factors are MillCaliber heuristics. They rank materials by how hard they are to cut. They are not measured specific cutting forces.
- Grades, species and tools vary widely. Your tool maker's published data takes priority over these baselines.
3. Diameter, machine and operation scaling
Intended chip = base chip × (D ÷ 0.25)^0.85 × machine factor × operation factor| Factor | Value | Rationale |
|---|---|---|
| Diameter exponent | 0.85 | Published charts give larger cutters larger chip loads, but less than in direct proportion to diameter. The 0.85 exponent is a MillCaliber heuristic. It has not been fitted to a dataset, and that validation is pending. |
| Machine factor: desktop / belt-driven | 0.55 | This reduces chip load (and cutting force) on less stiff machines. The values are planning heuristics, not measured stiffness. |
| Machine factor: semi-rigid benchtop | 0.75 | |
| Machine factor: rigid / ball-screw | 1.00 | |
| Operation: adaptive / pocketing | 1.00 | Default WOC 50% of D. This is the baseline condition. |
| Operation: slotting | 0.75 | Default WOC 100% of D. Full-width cuts have the most engagement and the hardest chip evacuation, so feed is reduced by 25% (heuristic). |
| Operation: finishing | 1.00 | Default WOC 10% of D. The light stepover is handled by chip-thinning compensation (section 4). |
Choosing an operation pre-fills WOC with its default. If you then edit WOC, the operation's feed factor still applies. The calculator adds a note when WOC and operation disagree, for example a full-width cut with Adaptive selected.
4. Radial chip thinning
When the radial width of cut ae is less than half the diameter, each tooth leaves the material before it reaches full chip thickness. The cutter engages over an angle φ where cos φ = 1 − 2·ae/D, and the maximum chip thickness is hmax = IPT × sin φ. Rearranging gives the radial chip-thinning factor (RCTF):
- This is a geometric derivation from the cutter's engagement angle, not a value copied from a source.
- Cap at 2.5×. The cap stops very small stepovers from producing extreme programmed feeds. It applies once WOC falls below about 4.2% of D. Beyond that point the actual chip is thinner than intended, which risks rubbing, and the calculator adds a note saying so.
- Axial chip thinning (ball-nose and chamfer tools) and lead-angle effects are not modeled.
5. RPM rounding and machine limits
- Rounding: ideal RPM is rounded down to the nearest 500 RPM. Rounding down keeps surface speed at or below the target, and many router spindles and VFDs are set in coarse steps.
- Spindle range: the result is then clamped to your min and max RPM.
- If it is capped at max, the surface speed is below target. Feed per tooth is kept, so the feed is lower.
- If it is raised to min, the surface speed is above target, which can add heat and tool wear. The calculator warns you.
- Feed limit: if
RPM × z × IPTexceeds your max cutting feed, RPM is lowered to keep feed per tooth:RPM = ⌊max feed ÷ (z × IPT) ÷ 100⌋ × 100. - Below min RPM (constrained estimate): if that RPM would fall below your min RPM, the spindle runs at min RPM and the feed is held at your max. Your limits are never overridden, so the intended chip thickness cannot be reached: the programmed feed per tooth becomes
max feed ÷ (min RPM × z)and the actual maximum chip is that value ÷ RCTF. The result is labeled Constrained estimate, and both the actual and intended chip are shown. Options are a higher max feed (if the machine can safely run faster), a lower min RPM, or fewer flutes. A smaller WOC lowers engagement but makes the chip even thinner, because the capped feed cannot rise to compensate. A larger WOC (up to 50% of D) thickens it, at the cost of more engagement. - Default limits: the defaults are generic values for each machine class, not data for any specific machine. Replace them with your machine's real limits.
- Desktop: 10,000–24,000 RPM, 200 in/min.
- Semi-rigid: 6,000–24,000 RPM, 150 in/min.
- Rigid: 6,000–24,000 RPM, 400 in/min.
6. Plunge feed
Plunge = 0.30 × cutting feed. This is a MillCaliber convention, not a tool-maker value, and it is unvalidated. One general source (MIT feeds-and-speeds notes) recommends keeping plunge below 50% of the feed rate. That supports plunging slower than cutting, but it does not establish 30% as a suitable setting for any particular tool. Whether a cutter can plunge at all, and how fast, depends on its geometry (center-cutting or not), material and the tool maker's guidance, which always takes priority. Square end mills are weak in straight plunges, and many router bits are not center-cutting. Use a ramp or helix entry when your CAM supports it, and check that your cutter can plunge at all.
7. Rigidity and the engagement guideline
Engagement = (WOC ÷ D) × (DOC ÷ D) × material factor ÷ rigidity factor Rigidity factor: desktop 1.0 · semi-rigid 1.6 · rigid 2.5| Engagement | Shown as |
|---|---|
| ≤ 0.55 | Moderate engagement |
| 0.55 – 0.90 | High engagement (amber) |
| > 0.90, or DOC > 3 × D | Exceeds engagement guideline (red) |
These thresholds are MillCaliber heuristics for flagging aggressive setups. They don't predict tool breakage or chatter, and they don't consider spindle power. A “moderate” result can still fail on a worn tool, long stick-out or poor workholding.
Separately, a note appears when DOC is more than 2 × D, so you can confirm flute length and stick-out.
8. Input validation
When any input is invalid, no values are shown and the copy button is disabled. Each field shows its own error.
| Input | Accepted range |
|---|---|
| Custom diameter | 0.03–1 in (0.8–25.4 mm) |
| Depth of cut (DOC) | 1% of D up to 4 × D |
| Width of cut (WOC) | 1% of D up to 100% of D |
| Min / max spindle RPM | Whole numbers 500–60,000, min lower than max |
| Max cutting feed | 1–2,000 in/min (25–50,800 mm/min) |
Blank, non-numeric, zero and negative values are rejected.
9. What each output means
- Spindle speed: the RPM after rounding and machine limits.
- Cutting feed: XY feed after any feed-limit adjustment.
- Plunge feed: 30% of cutting feed.
- Feed per tooth: the value to program, including chip-thinning compensation. “Actual max chip” is what the tooth really cuts at your stepover.
- Surface speed: the result at the actual RPM. The target is shown beneath it.
- MRR: material removal rate.
- Step over / step down: WOC as % of D, and DOC as a multiple of D.
Baseline vs actual: the material table shows baselines for a 1/4" cutter on a rigid machine. The result cards show actual values after every adjustment. “How this estimate was built” under the results lists each step from baseline to actual for your current inputs.
The copied “CAM values” are plain text for typing into your CAM tool's speed and feed fields, with units. They are not a G-code program.
Rounding: calculations run at full precision. Displayed and copied values use the same rule, round half up: feeds to 0.1 in/min or 1 mm/min, feed per tooth to 0.00001 in or 0.0001 mm, surface speed to 1 SFM or 1 m/min. The copied text always matches the screen.
10. Limitations
- Assumed tool: solid-carbide, sharp, square-end, with flute length longer than the DOC, short stick-out and low runout.
- Not modeled:
- coatings, helix angle, rake and edge prep, or brand-specific tool data;
- high-speed steel tools;
- ball-nose effective diameter and axial chip thinning;
- spindle power and torque curves;
- coolant and lubrication;
- workholding, tool deflection and runout;
- chip evacuation in deep slots;
- machine acceleration (short moves may never reach the programmed feed).
- Narrow coverage: five generic materials, and three machine classes described by a single factor each.
- Unvalidated values: the brass surface-speed target is above the one published example checked, and the aluminum target has not been confirmed for MillCaliber's assumed tooling. The diameter exponent, machine factors, operation factors and engagement thresholds are heuristics without a published basis. None of these has been validated against measured cuts. See the validation status.
- No professional review yet: no independent machining professional has reviewed the method so far.
Found a value that didn't work? Please email contact@millcaliber.com with your machine, cutter, material and what happened.
11. Validation status
This table shows the evidence found so far for each assumption. “Supported” or “within range” means published examples agree with the value. It does not mean the value has been tested on a machine or is suitable for your tooling. MillCaliber's calculation tests check that the site matches its own reference model; they do not show that the resulting settings are suitable for machining. No physical cutting tests have been run, and no independent machining professional has reviewed this method.
| Assumption | MillCaliber value | Evidence found | Status |
|---|---|---|---|
| RPM, SFM, feed and MRR equations | Standard formulas | Harvey Tool general machining guidelines | Supported (standard relationships) |
| Radial chip-thinning factor | D ÷ (2√(D·ae − ae²)), capped 2.5× | Geometric derivation (section 4); the cap value is a MillCaliber choice | Formula derived; cap unvalidated |
| Aluminum surface speed | 1,200 SFM | Published examples from 600 SFM (Amana, Productivity, UPenn) through 720–1,080 SFM (Helical) to about 3,900 SFM at 60,000 RPM (RobbJack, rigid setup) | Unvalidated: within the published range, but not confirmed for router-class cutters on desktop machines |
| Aluminum chip load, 1/4" | 0.0023 in/tooth | 0.0015–0.0024 (Productivity, slot to side rough), 0.002 (UPenn), 0.003–0.005 (CNC Router Store) | Within the published range |
| Wood and acrylic chip loads, 1/4" | 0.0060 / 0.0045 / 0.0040 | CNC Router Store chart (higher) | Below the cited example; suitability remains unvalidated |
| Wood and acrylic surface speeds | 1,800 / 1,400 / 900 SFM | No published source checked yet | Unvalidated |
| Brass surface speed | 600 SFM | 300 SFM (UPenn) | Unvalidated: above published example |
| Brass chip load, 1/4" | 0.0015 in/tooth | 0.002 (UPenn) | Below the cited example; suitability remains unvalidated |
| Diameter scaling exponent | 0.85 | Published charts imply about 1.0 for aluminum (chip load doubles from 1/4" to 1/2": Productivity, UPenn) and about 0.6 for brass (UPenn) | Lies between published examples; not fitted; unvalidated |
| Machine factors | 0.55 / 0.75 / 1.00 | No published source | Heuristic: unvalidated |
| Operation factors | Slotting 0.75 | Productivity's chart uses a lower chip load for slotting than for side milling (0.0015 vs 0.0024 at 1/4"), consistent in direction | Direction supported; size unvalidated |
| Engagement thresholds and factors | 0.55 / 0.90; K and rigidity values | No published source | Heuristic: unvalidated |
| Plunge feed | 30% of cutting feed | MIT feeds-and-speeds notes recommend less than 50% of feed rate (general guidance, not tool-specific) | Unvalidated: within that general rule of thumb, but the rule does not validate 30% for any specific tool; tool suitability and maker guidance apply |
Still needed before any value can be called validated
- Independent professional review of the method and constants by a qualified machinist or manufacturing engineer.
- Physical cutting tests on at least one machine per class (desktop, semi-rigid, rigid). For each material, run 1/8", 1/4" and 1/2" cutters at the calculated values. Record spindle load or sound, chatter, surface finish, chip form, tool temperature and wear, and compare with the predicted engagement level.
- Tool-maker data for the specific cutter types MillCaliber assumes, especially aluminum and brass surface speeds and chip loads by diameter. This is needed to replace or confirm the unvalidated values.
Until then, every output is a starting estimate. Your tool maker's published data takes priority.
12. Comparison sources
These sources were used to sanity-check equations and the order of magnitude of baselines. They do not endorse MillCaliber. Accessed 18 October – 6 November 2026.
- Harvey Tool, General Machining Guidelines: the RPM, SFM and IPM equations.
- CNC Router Store, CNC chip load chart: feeds & speeds by material: router chip-load ranges for a 1/4" bit.
- Amana Tool via ToolsToday, Aluminum cutting 55° helix speed chart (PDF): 1/4" 2-flute slot at 9,200 RPM, 74 in/min, 0.004 in/tooth, 600 SFM.
- Productivity Inc., RedLine 5-flute aluminum end mill speeds & feeds (PDF): 6061/7075/2024, 600 SFM slotting, 800–1,000 SFM side roughing, 650 SFM finishing; 1/4" chip loads 0.0015–0.005 by operation, doubling at 1/2".
- University of Pennsylvania MEAM design lab, Feeds and speeds guide (carbide end mills): aluminum 600+ SFM with 0.001/0.002/0.004 in/tooth at 1/8"/1/4"/1/2"; brass 300 SFM with 0.001/0.002/0.003 in/tooth.
- Helical Solutions, aluminum end mill speed & feed parameters (PDF): 720–1,080 SFM for its 3-flute aluminum end mills.
- RobbJack, Speeds & feeds for 6061 aluminum: A1-201 2-flute at 60,000 RPM, rigid setup; 0.0064 in/tooth and about 3,900 SFM at 1/4".
- MIT Center for Bits and Atoms, Feeds and speeds: plunge rate: plunge rate should be less than 50% of the feed rate.