Work-hardening materials — austenitic stainless (304/316), nickel superalloys (Inconel 718), and Hadfield manganese steel — harden under the cut because plastic deformation ahead of the edge raises surface hardness by 50-100% (and up to a typical 2-3x in Hadfield) before the next pass arrives. The single rule that prevents tool failure is to keep the chip thicker than the work-hardened layer (typically ~0.02-0.05 mm deep): use sharp positive geometry, an adequate depth of cut that reaches below the hardened skin, no dwelling, and climb milling so each edge enters fresh material rather than glazing across hardened surface. Feed and depth of cut that are too light — not too heavy — are the dominant cause of premature failure in these materials.
These materials share one failure mechanism — strain hardening at the cut surface — but each has a different sensitivity. This article is the cross-material deep dive on the mechanism and the universal tactics that apply to all of them. It deliberately does not repeat the per-grade speed/feed tables already published: for the full 304/316 parameter set see the stainless steel machining guide, for Ti-6Al-4V see the titanium alloy machining parameters guide, and for the carbide grade and binder logic see the carbide grade selection guide. A broader survey of difficult-material families is collected in the material machining complete guide. For the most extreme case in this family — nickel superalloys — see the Inconel superalloy machining guide; for the two-phase variant, see the duplex stainless machining guide.
Quick Work-Hardening Machining Reference
| Problem / Goal | Primary Action | Expected Impact |
|---|---|---|
| Tool glazes/rubs and fails fast in stainless or Inconel | Raise feed per tooth above the work-hardened-layer depth (≥0.05 mm/tooth for Inconel; ≥0.03 mm/tooth for 304) | Cutting replaces rubbing once chip thickness clears the ~0.02-0.05 mm hardened skin |
| Surface "glaze" / hardened skin after a stalled pass | Increase axial DOC so the edge reaches below the prior pass's hardened layer (typically ≥0.5 mm in Inconel) | Edge cuts fresh material instead of skidding on hardened surface |
| Notch wear at the depth-of-cut line | Vary radial DOC (ae) by 0.2-0.5 mm between passes; chamfer the entry | Wear distributes along the edge instead of concentrating at the air/skin boundary |
| Hadfield manganese steel hardens faster than it cuts | Use ceramic or heavy positive carbide, large DOC, single deep pass — avoid light skim cuts | Removing material below the hardened zone in one pass typically cuts cycle time and tool wear |
| Tool dwells at end of pass and leaves a hard spot | Program continuous retract; never pause spindle-engaged at depth | Eliminates the localized work-hardened patch that breaks the next edge |
Why These Materials Harden Under the Cut
Work hardening (strain hardening) occurs when plastic deformation ahead of and beneath the cutting edge increases the dislocation density of the surface layer, raising its hardness by roughly 50-100% in austenitic stainless and nickel alloys — and by a typical 2-3x in Hadfield manganese steel — before the next tooth or pass engages it. The mechanism is metallurgical, not a tooling defect, which is why parameter strategy matters more than tool brand.
In a face-centred-cubic (FCC) austenitic structure, dislocations cannot cross-slip easily, so deformation piles them up rather than relieving them. The thin layer the tool plastically smears — typically 0.02-0.05 mm deep under normal cutting — emerges harder than the bulk. The work-hardened layer in austenitic stainless can reach 250-300 HB after a single pass, roughly double the 150-185 HB annealed bulk hardness, a value cross-referenced with the detailed treatment in the stainless guide. In Inconel 718 the as-cut surface can rise from ~36-44 HRC bulk toward localized values 20-50% higher in the deformed skin.
Hadfield (austenitic manganese) steel is the extreme case: its surface can transform from ~200 HB to over 500 HB under impact or heavy deformation, which is precisely the property that makes it useful in rock crushers and rail crossings — and miserable to machine. The same FCC-stabilising manganese (11-14%) that resists abrasion in service also resists the cutting edge.
The practical consequence is counter-intuitive: light cuts make work-hardening materials harder to machine, not easier, because a shallow chip leaves the edge skidding across the layer the previous pass already hardened.
Glazing and Rubbing: The Core Failure Mode
Glazing (also called rubbing or burnishing) happens when chip thickness falls below the depth of the work-hardened layer, so the edge plows and smears the hardened skin instead of shearing through it — generating heat, accelerating flank wear by a typical 3-5x, and leaving a bright, hardened surface that defeats the next pass. This is the dominant way work-hardening materials destroy tools, and it is a parameter problem, not a speed problem.
When an edge rubs rather than cuts, three things compound:
- Heat with no chip to carry it away. A proper chip removes most cutting heat. A rubbing edge converts the energy to friction at the interface, raising local temperature into the regime where coatings oxidise and carbide softens.
- Self-reinforcing hardening. The rubbing action work-hardens the surface further, so each subsequent pass faces an even harder skin — a runaway loop that ends in rapid edge breakdown.
- Dimensional drift. As the glazed edge wears, it deflects and the part trends oversize on external features, with a finish that is bright and irregular rather than uniformly cut.
The diagnostic signature is distinctive: a glazed (mirror-bright) workpiece surface plus rapid, uneven flank wear is the fingerprint of rubbing below the work-hardened layer, as opposed to the matte, ragged surface that signals built-up edge. Both can occur together when feed is far too low.
Light Skim Cuts Are the Trap
The instinct when a work-hardening material misbehaves is to back off feed and depth "to be gentle." In austenitic stainless, Inconel, and especially Hadfield steel this does the opposite: it drops chip thickness below the hardened-layer depth, converts cutting into rubbing, and can accelerate flank wear by 3-5x in typical shop experience. If a pass is struggling, increase chip load and depth to get below the skin — do not reduce it.
The Universal Toolkit
Across all work-hardening materials, four parameter rules dominate: sharp positive geometry, an adequate depth of cut that reaches below the hardened layer, no dwelling, and climb milling — together they keep every edge cutting fresh material instead of glazing across hardened surface. These apply regardless of grade; the per-material numbers differ, but the logic is identical.
Sharp, positive rake geometry. A keen positive edge shears the layer cleanly at lower cutting force, reducing the deformation that creates the hardened skin in the first place. Honed or heavily negative edges plow more material ahead of the cut and deepen work hardening. PVD-coated edges (typically 5-15 µm edge radius) preserve sharpness; thick CVD coatings round the edge (25-40 µm) and are generally avoided for these materials except in stable continuous turning. Positive geometry is preferred for work-hardening alloys because lower cutting force means a thinner deformed layer for the next pass to handle.
Adequate depth of cut — get under the skin. The axial and radial DOC must place the edge below the layer the previous operation hardened. As a working rule, the depth of cut should exceed the work-hardened layer depth (typically 0.02-0.05 mm) by a comfortable margin — in Inconel that often means ap ≥ 0.5 mm even in finishing. A finishing pass that removes less than the hardened skin will ride on top of it and glaze.
No dwelling. A tool paused at depth with the spindle turning work-hardens a localized spot that breaks the next edge. Never dwell at depth in a work-hardening material — program continuous feed and a positive retract, and avoid spring passes that re-engage the hardened surface at near-zero chip load.
Climb milling. In climb (down) milling each tooth enters at maximum chip thickness and exits at zero, so it engages fresh material immediately rather than skidding into the hardened surface left by the prior tooth. Conventional milling reverses the chip order and systematically deepens work hardening in austenitic alloys, which is why climb milling is the baseline assumption for these materials.
Match Feed and DOC to the Hardened Layer, Then Set Speed
The first decision in a work-hardening material is feed and depth — both must clear the ~0.02-0.05 mm hardened layer with margin — not speed. Set chip load and DOC to guarantee cutting (not rubbing) first; tune cutting speed second to control interface temperature. Reversing that order is the most common reason tools glaze in stainless and Inconel.
How the Three Material Families Differ
The families differ mainly in hardened-layer severity and the dominant secondary wear mode: austenitic stainless is adhesion-driven (built-up edge), nickel superalloys are heat-and-abrasion driven (notch and crater), and Hadfield steel hardens so aggressively that bulk removal in one deep pass beats any multi-pass strategy. The toolkit is shared; the emphasis shifts.
Austenitic stainless (304/316). Moderate work-hardening, with built-up edge as the primary secondary failure. The hardened layer is ~0.02-0.03 mm; feed per tooth above ~0.03 mm (304) clears it. Detailed speeds, feeds, and coating choices are in the stainless steel machining guide — this article does not repeat them.
Nickel superalloys (Inconel 718). The most demanding mainstream case. Inconel 718 retains roughly 80% of its room-temperature strength at 650°C, so the cutting zone stays hot and notch wear at the depth-of-cut line is the dominant failure mode. Run low speed (typically 20-40 m/min with carbide; higher with ceramics under continuous cuts), keep DOC well below the surface but deep enough to clear the hardened skin, vary ae to spread notch wear, and chamfer or ramp the entry. AlCrN coatings are preferred for Inconel finishing because their oxidation onset near 1,100°C exceeds TiAlN's ~800°C limit, giving thermal headroom in the hot cutting zone.
Hadfield manganese steel. Because Hadfield steel can more than double its surface hardness under deformation, the winning strategy is a single deep pass with a heavy positive edge or ceramic — large ap, low-to-moderate speed, generous chip load — rather than multiple light passes that each re-harden the surface. Many shops machine Hadfield only in the as-cast (softer) condition before the work-hardening service treatment, or accept ceramic tooling and reduced metal-removal rates. Grinding is sometimes substituted where geometry allows.
Diagnosing What Went Wrong
Read the workpiece and the edge together: a bright glazed surface with fast flank wear means the chip was thinner than the hardened layer (rubbing); a notch at the depth-of-cut line means radial engagement stayed constant; built-up edge with a ragged finish means feed was marginally low; uniform flank wear means the parameters are correct. Each pattern points to a specific fix, so diagnose before re-cutting.
- Glazed/burnished surface + uneven flank wear → chip thickness below the hardened layer. Raise feed and DOC to get under the skin.
- Notch at the ae line → engagement too constant. Vary radial DOC 0.2-0.5 mm between passes; chamfer the entry.
- Ragged matte finish, dimensional drift → built-up edge. Raise feed per tooth, confirm climb milling, use a sharp PVD edge (see the stainless guide for grade-specific floors).
- Sudden edge fracture after a stall → a dwell or spring pass hardened a spot. Eliminate dwelling and zero-chip-load re-engagement.
Industry practice replaces the edge when uniform flank wear reaches the ISO 3685 criterion (VB = 0.3 mm average for finishing, 0.6 mm maximum for roughing); in work-hardening materials, applying the tighter 0.3 mm criterion is conservative because rubbing can cascade once the edge dulls. ISO 513 class M is the reference work-material group for austenitic stainless and class S for heat-resistant superalloys, which is why grade selection follows those groups rather than the P (steel) tables.
Summary
Keep the chip thicker than the hardened layer: sharp positive geometry, adequate DOC under the skin, no dwelling, climb mill.
Work-hardening materials fail when the edge rubs instead of cuts. Keep feed and depth of cut above the ~0.02-0.05 mm work-hardened layer so every pass shears fresh metal, use sharp positive geometry to minimise the deformed skin, never dwell at depth, and climb mill so each edge enters at maximum chip thickness. Austenitic stainless adds built-up edge as a secondary mode, Inconel 718 adds heat-driven notch wear (use low speed and AlCrN), and Hadfield steel rewards a single deep pass over light skims. For per-grade speeds and feeds, defer to the dedicated stainless, titanium, and carbide-grade guides — this article covers the shared mechanism and the universal toolkit that applies across all of them.
Why do light cuts make work-hardening materials harder to machine?
Light cuts drop chip thickness below the ~0.02-0.05 mm work-hardened layer, so the edge rubs and smears the hardened skin instead of shearing through it. This generates heat with no chip to carry it away and can accelerate flank wear by 3-5x. Increase feed and depth of cut to get below the hardened layer.
What is glazing in machining and how do I prevent it?
Glazing is when the cutting edge burnishes a work-hardened surface instead of cutting it, leaving a bright, hardened skin and rapid uneven wear. Prevent it with sharp positive geometry, a depth of cut that reaches below the ~0.02-0.05 mm hardened layer, no dwelling at depth, and climb milling so each edge enters fresh material.
How much harder does austenitic stainless get when work-hardened?
The surface layer of 304/316 stainless can reach 250-300 HB after a single cutting pass, roughly double the 150-185 HB annealed bulk hardness — about a 50-100% increase. The hardened layer is typically 0.02-0.03 mm deep, which sets the minimum chip thickness needed to keep cutting rather than rubbing.
Why is Hadfield manganese steel so difficult to machine?
Hadfield steel work-hardens from about 200 HB to over 500 HB (up to ~2-3x) under deformation, the same property that makes it abrasion-resistant in service. Light passes re-harden the surface each time, so the effective strategy is a single deep pass with a heavy positive carbide or ceramic edge, or machining in the softer as-cast condition.
Should I climb mill or conventional mill work-hardening materials?
Climb (down) milling is preferred because each tooth enters at maximum chip thickness and exits at zero, engaging fresh material immediately. Conventional milling reverses the chip order so the tooth skids into the previously hardened surface, deepening work hardening and accelerating built-up edge in austenitic stainless and nickel alloys.
Sources
- ISO 513: Classification and Application of Hard Cutting Materials
- ISO 3685: Tool Life Testing with Single-Point Turning Tools
- Sandvik Coromant: Machining Heat-Resistant Super Alloys (HRSA)
- Kennametal: Work Hardening and Difficult-to-Machine Materials
- Machinery's Handbook, 31st Edition — Machining Stainless Steels and Superalloys
