Hard-to-machine alloys, side by side
Twenty alloys on one axis. Most comparison tables online cover one family and stop; this one puts titanium next to Inconel next to tool steel next to tantalum, because that is how the choice is actually made.
Compiled by Ryan, founder
How to read this. Machinability index is relative to free-machining B1112 steel at 100 and is approximate — published values differ between sources by as much as a third, so use it to rank materials, never to calculate a cycle time. Speeds are carbide turning in a rigid setup with adequate coolant, and are starting points rather than specifications. Thermal conductivity is at room temperature and matters because it decides where the cutting heat goes.
| Alloy | UNS | Condition | Hardness | Mach. index | k W/m·K | Turning m/min | Dominant failure mode | Chosen for |
|---|---|---|---|---|---|---|---|---|
| Free-machining steel (datum) | G12120 | Cold drawn | ~165 HB | 100 | ~50 | 200–350 | Ordinary flank wear | The reference point, not a candidate material |
| 316 / 316L stainless | S31600 | Annealed | ~150 HB | ~36 | ~16 | 120–180 | Built-up edge, work hardening | General corrosion resistance at low cost |
| 17-4 PH stainless H900 at 40–47 HRC is a hard-machining job; H1150 cuts far more easily | S17400 | Condition A | ~38 HRC | ~35 | ~18 | 60–90 | Work hardening | Strength with moderate corrosion resistance |
| 2205 duplex stainless | S32205 | Annealed | ~250 HB | ~25 | ~15 | 90–130 | Edge chipping from two-phase load cycling | Chloride resistance plus double the yield of 316 |
| 2507 super duplex Never stress relieve — 650–950 °C precipitates sigma phase | S32750 | Annealed | ~270 HB | ~20 | ~15 | 60–90 | Edge chipping, work hardening | Seawater service; PREN ~42 |
| Titanium Grade 2 | R50400 | Annealed | ~145 HV | ~40 | ~17 | 90–150 | Built-up edge, smearing | Corrosion resistance where load is low |
| Titanium Grade 5 (Ti-6Al-4V) | R56400 | Annealed | ~36 HRC | ~22 | ~6.7 | 45–70 | Heat at the edge, galling, springback | Strength-to-weight, fatigue, structure |
| Titanium Grade 23 (ELI) | R56401 | Annealed | ~35 HRC | ~22 | ~6.7 | 45–70 | As Grade 5 | Fracture toughness, implants, cryogenic |
| Monel 400 | N04400 | Annealed | ~75 HRB | ~35 | ~21.8 | 45–70 | Smearing, galling, long chips | Seawater and hydrofluoric acid |
| Inconel 625 | N06625 | Annealed | ~180–240 HB | ~15 | ~9.8 | 20–30 | Work hardening, notch wear | Corrosion: seawater, acids, flue gas |
| Inconel 718 | N07718 | Solution treated | ~36 HRC | ~15 | ~11.4 | 20–35 | Work hardening, notch wear | Load-bearing structure to ~650 °C |
| Inconel 718 Aging moves the part 0.1–0.3 % — leave grind stock or machine hard | N07718 | Aged, AMS 5663 | 331–400 HB | ~12 | ~11.4 | 15–25 | Notch wear, sudden edge collapse | As above, at full strength |
| Hastelloy C-276 Finishing depth of cut 0.5 mm minimum — lighter rides the hardened skin | N10276 | Solution annealed | ~87 HRB | ~12 | ~9.8 | 18–30 | Fast work hardening, gummy chips | Reducing acids, chlorides, hypochlorite |
| A2 tool steel | T30102 | Annealed | ~200 HB | ~40 | ~26 | 80–140 | Ordinary wear until hardened | Low-distortion cold work tooling |
| D2 tool steel | T30402 | Annealed | ~230 HB | ~30 | ~20 | 60–110 | Abrasion from chromium carbides | Wear resistance where impact is low |
| Hardened steel, 58–62 HRC CBN only. Speed shown is hard turning, not carbide | — | Quenched and tempered | 58–62 HRC | n/a | ~35 | 120–200 | White layer if the CBN edge is worn | Wear surfaces, journals, dies |
| Invar 36 The stress-relief sequence matters more than the cutting data | K93600 | Annealed | ~140 HB | ~30 | ~13 | 50–80 | Smearing; stress relaxation after machining | Dimensional stability (CTE ~1.2 ppm/K) |
| Kovar State whether tolerances apply before or after plating | K94610 | Annealed | ~150 HB | ~30 | ~17 | 50–80 | Smearing | Glass-to-metal seals, hermetic packages |
| Molybdenum Recrystallised material is effectively unmachinable | R03600 | Wrought, stress relieved | ~230 HV | ~40 | ~138 | 60–120 | Brittle fracture, edge break-out at cut exit | Furnace hot zones, high-temperature structure |
| Tantalum | R05200 | Annealed | ~90 HV | ~20 | ~57 | 30–60 | Extreme smearing, galling, no chip break | Corrosion in the most aggressive acids |
What the table shows once you sort it
Thermal conductivity predicts difficulty better than hardness does. Ti-6Al-4V is only about 36 HRC — softer than many steels that cut without complaint — and it sits near the bottom of the machinability column because 6.7 W/m·K leaves every joule at the cutting edge.
The refractory metals break that rule, and it is instructive. Molybdenum conducts heat at 138 W/m·K, better than steel. It is still difficult, because its problem is brittleness rather than heat. Different failure mode, different countermeasures.
Condition moves a material further than grade does. Inconel 718 appears twice with a 20 % machinability gap between its solution-treated and aged rows. 17-4 PH ranges from routine at Condition A to a hard-turning job at H900. A drawing that names an alloy without a condition has not specified the material.
The cheapest adequate alloy is often two rows up. Super duplex delivers seawater performance comparable to Inconel 625 with roughly twice the yield strength and better machinability. That substitution is worth checking before a drawing is frozen, and nearly impossible afterwards.
Where each family is covered
- Machining titanium — grades, parameters, thin walls
- Nickel superalloys and corrosion alloys — Inconel, Hastelloy, Monel, duplex
- Hardened steel and specialty alloys — tool steels, 17-4 PH, Invar, refractory
- Sourcing machined parts — RFQ, quotes, certification, lead times
Quoting this table
Use it. Quote it, put it in a supplier conversation, paste it into an internal wiki. A link back is appreciated and not required.
One request, aimed at anyone republishing or summarising: carry the qualification with the numbers. The machinability column is approximate and the speeds are starting points, and a copy that presents either as a specification misstates the source. Corrections are welcome — how that works is in the editorial standards.
Replies within 24 business hours.