Explainers · 2026-07-03 · ~3,900 words

Patreon for blacksmithing creators: forge temperature physics, steel hardenability mechanics, quench media chemistry, forge welding flux science, grain refinement, and the Apple Tax

Blacksmithing and bladesmithing Patreon documentation retains patrons when it explains the physics and chemistry underneath the visible process — why the steel changes color as it heats, how carbon content sets an absolute ceiling on achievable hardness, what quench oil chemistry does at the surface of hot steel, why borax makes forge welding possible at all, and how grain structure at the moment of quench determines blade toughness months later. This post covers the deep technical layer that the SEO overview and YouTube tutorials structurally cannot: the underlying mechanisms, with values, units, and documentation protocols.

The physics of forge temperature color

Blackbody radiation as the emission mechanism

When a creator documents “orange heat” or “lemon-yellow,” they are describing the visible output of a physical process called blackbody radiation. Any object above absolute zero emits electromagnetic radiation, and the spectrum of that radiation — the distribution of power across wavelengths — is determined by the object’s temperature alone. As temperature rises, the peak emission wavelength shifts toward shorter wavelengths: at moderate temperatures the peak is deep in the infrared and the visible output is negligible; as temperature climbs through 400°C (752°F), emission in the deep red end of the visible spectrum becomes detectable; at 1000°C (1832°F), the spectral peak has shifted enough that orange wavelengths dominate the visible output; at 1300°C (2372°F), the peak is closer to yellow-green and the visible emission is white-yellow.

Steel in the forge behaves as a near-blackbody emitter — the spectral emission is governed by temperature rather than by steel grade, carbon content, or alloy additions. This means the color you observe is a reliable measure of temperature, with one important caveat: you are observing the surface temperature, not the core temperature. The surface of a thick billet heats faster than the core, and color observation during rapid heating in a high-output forge can mislead: the exterior surface may be at lemon-yellow while the interior is still at bright red. For thin stock and blade profiles, the thermal mass is small enough that surface and core temperatures equalize quickly. For thick cross-sections — hammers, tools, heavy forgings — a soak at heat is required to allow the core to reach the surface temperature before manipulation.

The temperature-to-color mapping

The following values apply in normal indoor forge light with a dark background. Bright ambient light washes out the low-temperature glows; outdoors in direct sunlight, the first visible glow does not appear until substantially higher temperature than the values below suggest. Document the observation conditions in forge records.

Black heat: 120–200°C (248–392°F). No visible emission. Hot enough to cause severe skin burns with brief contact; handle as if red hot. First visible glow: approximately 400°C (752°F), dull red at the limit of dark-adapted vision. Blood red: 530–580°C (986–1076°F); the color of a deep arterial red at the limits of photopic vision. Dark cherry red: 700–750°C (1292–1382°F); close to the Curie transition. Full cherry red: 815–870°C (1500–1598°F); typical quench-heat-visible assessment range for some spring steels. Bright red: 870–980°C (1598–1796°F). Orange: 980–1050°C (1796–1922°F); general drawing and shaping heat for mild steel. Orange-yellow: 1050–1150°C (1922–2102°F); heavy drawing and punching heat; approaching forge welding range for high-carbon steels. Lemon-yellow: 1150–1260°C (2102–2300°F); forge welding range for mild steel, and the heat at which borax flux is fully reactive. White-yellow: 1260–1350°C (2300–2462°F); upper working limit before sparkling (burning). At or before sparkling, the steel surface is irreversibly damaged.

Forge atmosphere and apparent color shift

The forge atmosphere — the ratio of fuel to air — changes what you observe at a given temperature. In an oxidizing atmosphere (excess air relative to fuel), iron at the steel surface reacts with free oxygen rapidly to form iron oxide scale: primarily magnetite (Fe₃O₄) at lower temperatures and wustite (FeO) at temperatures above approximately 570°C (1058°F). Iron oxide is darker and absorbs more light than clean iron at the same temperature, and it emits at a slightly lower effective temperature than the bright iron underneath. The visible result is that the steel appears darker and cooler than its true temperature when heavily scaled in an oxidizing atmosphere. In a reducing atmosphere (fuel-rich, excess fuel relative to air), free oxygen is consumed by the combustion reaction and little remains to react with the steel surface. Scale formation is suppressed, the surface remains relatively clean, and the visible emission more accurately represents true metal temperature.

For documentation, note whether the forge was running oxidizing or reducing during each heat, particularly for observations near critical temperature or forge welding heat. A bladesmith who normalizes in a slightly oxidizing forge and quenches in a slightly reducing atmosphere produces different surface conditions than one who runs a consistent reducing fire throughout. These details are invisible in process videos but are the kind of parameters that explain why two creators using the same steel achieve different results at the same nominal temperature.

The Curie point: a reproducible calibration reference

Color-based temperature assessment depends on trained eye judgment and observation conditions. A physical calibration reference that does not depend on color is the Curie transition. Iron is ferromagnetic below its Curie temperature — the crystal structure contains magnetic domains that align and create a net magnetic moment, making iron strongly attracted to external magnets. At the Curie temperature, the thermal energy is sufficient to disorder these magnetic domains: the iron transitions from ferromagnetic to paramagnetic, losing its strong magnetic response essentially instantaneously. For pure iron, the Curie temperature is 769°C (1416°F). For steel with approximately 0.6–1.0% carbon content, the Curie transition occurs at approximately 723°C (1333°F) — the same temperature as the Ac1 eutectoid point, because both phenomena occur at the same crystal-structure-related temperature in high-carbon steel.

The practical technique: hold a rare-earth magnet (neodymium) close to the steel as it heats — not touching, because the magnet will magnetize scale particles and contaminate the surface, but close enough to feel the attraction. The magnet will pull visibly toward the steel at all temperatures below the Curie point. At the transition temperature, the attraction vanishes immediately and the magnet hangs free. The transition is sharp enough that a bladesmith can identify it to within 5–10°C by feeling the moment the pull disappears. For 1084 steel, the magnet-drop moment corresponds well to the Ac1 temperature needed for the beginning of austenite formation during heat treatment. Document the magnet-drop observation temperature (estimated from simultaneous color observation) and the forge output setting at that moment; over multiple heats this builds a calibrated relationship between forge settings, color, and Curie-drop temperature for the creator’s specific equipment.

Steel carbon content and hardenability mechanics

The iron-carbon phase diagram: crystal structures and critical temperatures

Iron is polymorphic — it exists in multiple crystal structures depending on temperature. Below 912°C (1674°F), iron is in the alpha phase, with a body-centered cubic (BCC) crystal structure: eight iron atoms at the corners of a cube, one at the center. The BCC lattice has octahedral and tetrahedral interstitial gaps between atoms, but they are small — the maximum carbon that can dissolve in BCC iron at room temperature is approximately 0.02% by weight, insufficient to significantly harden the metal. Above 912°C (1674°F), iron undergoes an allotropic transformation to the gamma phase — face-centered cubic (FCC) structure: eight atoms at the corners, one at the center of each face. The FCC arrangement creates larger interstitial gaps (particularly the octahedral sites at the center of the cube and midpoints of the edges), and carbon atoms fit comfortably in these positions. The FCC austenite phase can dissolve up to 2.1% carbon by weight in solid solution.

The Ac1 critical temperature — 723°C (1333°F) for the iron-carbon eutectoid — is the temperature at which the pearlite structure (alternating lamellar plates of ferrite and iron carbide) begins converting to austenite on heating. Below Ac1, pearlite and ferrite coexist in equilibrium. Above Ac1, austenite nucleates at the pearlite grain boundaries and grows inward as the temperature rises. The Ac3 temperature varies with carbon content: for a 0.4% carbon steel, Ac3 is approximately 810°C (1490°F); for 0.8% carbon (the eutectoid composition), Ac1 and Ac3 converge at 723°C. For hypereutectoid steels above 0.8% carbon (such as 1095 with approximately 0.95% C), there is a separate Accm temperature above which excess carbides fully dissolve into austenite. The bladesmith’s concern is austenitizing above Ac3 (or Accm for hypereutectoid steels) to achieve a fully austenitic structure before quenching.

Carbide dissolution, hold time, and undissolved carbide effects

Simply reaching austenitizing temperature is not sufficient to dissolve all carbides. Carbide dissolution is a diffusion-controlled process: carbon atoms must migrate from carbide particles through the iron lattice into the surrounding austenite, and this migration takes time proportional to the diffusion coefficient of carbon in austenite at that temperature (approximately 2 × 10−11 m²/s at 850°C, roughly doubling for every 50°C rise). Small, well-distributed carbides dissolve quickly; large carbide clusters from coarse pearlite or carbide precipitation at prior grain boundaries may require several minutes at temperature for full dissolution.

Undissolved carbides at quench time have two negative consequences. First, they reduce the carbon concentration in the surrounding austenite, lowering the carbon available to trap in the martensite lattice and therefore lowering the achievable hardness below the theoretical maximum for that steel’s nominal carbon content. Second, undissolved carbide particles in the final martensite microstructure can act as stress concentration points, reducing toughness. For 1084 steel at a typical austenitizing temperature of 800–815°C (1475–1500°F), a hold of 3–5 minutes at temperature after the blade reaches full heat is generally sufficient for complete carbide dissolution, given the fine, uniform pearlitic microstructure that proper normalization produces. Longer hold times at temperatures well above Ac3 risk austenite grain growth, which has its own negative effects on toughness.

Martensite formation: the diffusionless shear transformation

When austenitized steel is quenched faster than its critical cooling rate, the FCC austenite lattice cannot transform to the equilibrium BCC ferrite-plus-carbide structure because carbon diffusion — which is required for the carbide precipitation step — is thermally suppressed. Instead, the lattice undergoes a diffusionless shear transformation: entire planes of atoms shift position cooperatively, without any atom moving more than one atomic diameter, producing a new crystal structure almost instantaneously. This is martensite.

The martensite crystal structure is body-centered tetragonal (BCT) — a BCC-like structure in which one axis is elongated relative to the other two. The elongation is caused by the carbon atoms trapped in the interstitial positions they occupied in the austenite lattice: they cannot diffuse out during the diffusionless transformation, so they remain in their sites but the surrounding iron lattice has shifted from FCC to BCT around them. The BCT distortion — the unequal axis lengths — is directly proportional to the carbon content: higher carbon content produces more interstitial carbon, greater axial distortion, and a harder martensite.

The hardness relationship with carbon content is approximately: at 0.6% C, the maximum achievable hardness is approximately 63 HRC; at 0.8% C, approximately 65 HRC; at 1.0% C, approximately 67 HRC theoretical maximum. In practice, retained austenite — austenite that did not transform during quench because the martensite-finish temperature (Mf) of high-carbon steels is below room temperature — reduces achieved hardness below theoretical maximum. The martensite-start temperature (Ms) for 1084 steel is approximately 280°C (536°F); for 1095 it is approximately 260°C (500°F). The martensite-finish temperature (Mf) for 1095 is below room temperature, meaning that after a room-temperature quench some austenite remains untransformed. Tempering cycles and subzero quench treatments can address retained austenite if maximum hardness is the design goal.

Quench media chemistry and selection

Heat transfer rate and the critical cooling rate

The function of the quench is to extract heat from the austenitized steel faster than the critical cooling rate — the minimum cooling rate below which austenite begins to transform to pearlite or bainite rather than martensite. The critical cooling rate is a property of the steel (determined by its composition: alloying elements like chromium, molybdenum, and nickel significantly slow the pearlite/bainite formation kinetics and allow martensite formation at lower cooling rates), not a property of the quench medium alone. For simple high-carbon steels like 1084 and 1095, the critical cooling rate is relatively high — these steels are not deeply hardenable and require a fast quench to achieve martensite throughout the cross-section. For alloy steels like 5160, alloying elements slow the transformation kinetics enough that an oil quench is sufficient where 1084 might need a faster medium.

Water and brine: the Leidenfrost effect and its disruption

Water provides the highest heat transfer coefficient of common quench media, but it does so unevenly. When very hot steel (above approximately 300°C / 572°F) contacts water, the steel surface instantly vaporizes the adjacent water layer, forming a continuous vapor blanket around the steel surface. This is the Leidenfrost effect: the steel is surrounded by superheated steam rather than liquid water, and steam conducts heat far less efficiently than liquid water. The vapor blanket insulates the steel from direct liquid contact for the first fraction of a second, then collapses as the surface cools and vapor production slows, at which point liquid water contacts the steel and heat extraction accelerates dramatically. The result is an uneven, two-phase quench profile: slow initial cooling while the vapor blanket persists, then a violent transition to fast cooling when it collapses.

Brine — water saturated with sodium chloride at approximately 350 grams per liter — disrupts the Leidenfrost vapor blanket earlier than pure water. The mechanism is still under study, but the dominant explanation is that dissolved salt ions promote nucleate boiling at the steel surface rather than film boiling: instead of a continuous steam film, discrete steam bubbles form and collapse rapidly, maintaining liquid water contact with the steel surface from the beginning of the quench. The result is a slightly faster initial cooling rate than pure water and a less pronounced two-phase transition. Brine was historically used for shallow-hardening steels that required water-fast quench rates; for most modern high-carbon blade steels and oil-quench tools, brine is rarely the optimal choice because the already-high thermal shock of water quench is not compensated by the marginal cooling rate improvement.

Oil quench: Parks 50 vs canola and the flash point consideration

Oil quench media provide moderate cooling rates with substantially less thermal shock than water. Oil’s lower specific heat capacity and higher viscosity mean it extracts heat more slowly than water per unit area of contact, but it maintains liquid-phase contact with the steel surface from the beginning of the quench without the vapor blanket issues that afflict water. The result is a smoother, more uniform heat extraction profile.

Parks 50 is a petroleum-based quench oil with a flash point of approximately 177°C (350°F). Its cooling rate in the critical transformation range (approximately 550–300°C) is significantly faster than canola or motor oil but substantially less violent than water. For 1084 steel — which has a relatively high critical cooling rate for a simple carbon steel — Parks 50 is the standard choice because it achieves martensite consistently through thin to moderate cross-sections without the warp and crack risk of water. For 1095, which has slightly higher carbon and a slightly lower critical cooling rate, Parks 50 also works reliably.

Canola oil is a vegetable oil with a flash point of approximately 190–220°C (374–428°F) and a cooling rate substantially slower than Parks 50 in the critical transformation range. For simple geometries, thick cross-sections, and steels with generous transformation windows, canola may achieve full martensite transformation. For thin blade profiles and steels at the faster end of the critical cooling rate spectrum, canola may produce soft spots — zones where cooling was not fast enough to suppress bainite or pearlite formation. Document which oil was used, the oil temperature at quench time (oil at 50–60°C / 122–140°F quenches slightly faster and more uniformly than cold oil because pre-heated oil has lower viscosity and better wets the steel surface), and the Rockwell result to correlate quench medium choice with achieved hardness.

Flash point matters not as a quality indicator but as a safety parameter: the quench tank should be operated at a temperature well below the flash point of the oil in use. During a quench, the oil temperature at the steel surface spikes dramatically for the first second; if the bulk oil temperature is too close to the flash point, the localized surface temperature can reach flash point and the oil vapor above the tank can ignite. Keep the bulk oil temperature below 50% of the flash point as a safe operating margin: Parks 50 at flash point 177°C (350°F) should be operated below approximately 90°C (194°F).

Entry orientation and the mechanics of warp prevention

Warping during quench occurs because martensite occupies approximately 4% greater volume than the austenite from which it forms. The BCT lattice, with its elongated axis and trapped interstitial carbon, is physically larger per unit cell than the FCC austenite. Sections of the blade that transform at different times produce local volume expansion at different moments; the differential expansion across the blade geometry creates internal stresses that resolve as geometric distortion — the blade curves, twists, or develops an uneven profile.

Edge-down quench entry addresses this for most blade profiles. The blade enters the oil with the edge pointing down; the edge — being the thinnest section — has the smallest thermal mass and will cool fastest regardless of orientation. Entering edge-down means the edge contacts oil first and its quench begins first; the spine, entering last, has the edge’s transformation already underway when it begins its own cooling. The resulting differential is smaller than it would be if the spine entered first and completed its volume expansion before the edge began its transformation. For tip-heavy thin-profile blades, tip-first entry equalizes the cooling rate along the blade length: the tip, with minimal thermal mass, would quench almost instantaneously if it were last to enter the oil, while the ricasso would still be at austenite temperature — this differential along the length produces a lengthwise curve. Entering tip-first keeps the length in a consistent cooling sequence.

Document entry orientation, oil temperature, and the warp assessment (straight, minor curve, significant curve, twist) immediately after quench, before tempering. Straightening during tempering is possible when the blade is briefly reheated, but a blade with a major quench warp will retain residual stress even after straightening and tempering. The quench documentation record builds a dataset across multiple blades that reveals whether warp patterns are associated with specific entry techniques, specific blade geometries, or specific oil temperatures.

Forge welding flux mechanics

Iron oxide formation at forge welding temperature

At the temperatures required for forge welding — approximately 1250–1300°C (2282–2372°F) for mild steel — iron oxidizes with extraordinary speed. The primary iron oxide at these temperatures is wustite, FeO, which forms at the steel surface within seconds of exposure to oxygen in the forge atmosphere. Wustite is a non-stoichiometric compound: its actual iron-to-oxygen ratio is approximately Fe₀.₉₄O rather than the theoretical FeO, because iron vacancies in the crystal lattice are energetically favorable at these temperatures. At lower temperatures (below approximately 570°C / 1058°F), the stable oxide is magnetite (Fe₃O₄); at higher temperatures, wustite dominates.

Wustite on the steel surface prevents forge welding by interposing an oxide layer between the two metal faces that cannot form a metallic bond. A metallic weld requires iron atoms from both surfaces to be in direct contact close enough for their electron orbitals to overlap and form metallic bonds — this is impossible when both surfaces are coated with iron oxide. The historical solution to forge welding in open-fire conditions without flux — working extremely fast in a clean reducing fire and striking the weld in the first fraction of a second after the pieces leave the fire — was impractical for complex welds. Flux is the practical solution.

Borax chemistry: the iron borate slag mechanism

Borax (sodium tetraborate decahydrate, Na₂B₄O₇·10H₂O) melts at approximately 740°C (1364°F). Applied to steel that is at orange heat (approximately 900–1000°C / 1652–1832°F), the borax crystals contact the hot steel surface and immediately melt, forming a liquid glassy layer that flows across the metal surface by surface tension. At this stage, the liquid borax physically excludes atmospheric oxygen from the steel surface, suppressing further wustite formation on clean steel beneath it.

As the steel continues heating toward welding temperature, the molten borax reacts with any wustite already present on the steel surface: the boron oxide component of borax (B₂O₃) reacts with FeO to form iron borate compounds, primarily iron(II) borate phases including FeB₂O₄ and Fe₂B₄O₇. These compounds are glassy liquid slag at welding temperature, physically soluble in the borax melt, and most importantly mechanically weak — they can be squeezed out from between the weld surfaces by hammer pressure. The sequence at the moment of welding: the two steel surfaces, protected from fresh FeO formation by the liquid borax layer and cleaned of prior FeO by the iron borate reaction, are brought together at lemon-yellow to white-yellow heat. The first hammer blow applies compressive force that physically expels the liquid iron borate slag from between the surfaces outward, allowing direct iron-to-iron contact at the interface. If the temperatures, surfaces, and hammer blow are correct, the iron atoms from both surfaces are close enough and energetic enough to form metallic bonds — the weld sets.

Flux timing, the weld window, and sparkling as a diagnostic

Flux timing determines whether the iron borate reaction is complete before the weld attempt. Apply borax when the steel is at orange heat (approximately 950–1000°C / 1742–1832°F) — hot enough that the borax melts instantly on contact and flows across the surface, but cool enough that there is time remaining in the heat-up to welding temperature for the borax-FeO reaction to proceed. Applying flux at welding temperature means the borax has barely melted before the steel must leave the fire; the iron borate reaction is incomplete, and the weld surfaces are not properly protected. Applying flux at too low a temperature means the borax partially melts but may not flow uniformly across the surface, leaving unprotected zones.

The weld window is the temperature range above orange-yellow and below sparkling. Sparkling from the steel surface is a definitive diagnostic: it means the steel is burning — iron at the surface is reacting with oxygen so violently that the exothermic oxidation reaction ejects molten FeO particles as luminous sparks. Above the burning threshold, surface iron is being consumed, carbon at the surface is being oxidized and lost (decarburization in the extreme case, combustion in the burning case), and austenite grain size at the surface is growing uncontrollably. A weld attempted in burning steel will be mechanically compromised and full of slag inclusions from the excess iron oxide.

High-carbon steel welds at lower temperature than mild steel. The solidus temperature of steel — the temperature at which the surface begins to approach a semi-plastic or incipient liquid state — decreases as carbon content increases. For 1095 steel (approximately 0.95% carbon), the weld window is approximately 1100–1150°C (2012–2102°F) rather than 1250–1300°C (2282–2372°F) for mild steel. Attempting to weld high-carbon steel at mild-steel welding temperature risks burning the surface or approaching the solidus, where the steel begins to behave as a semi-liquid and the hammer blow produces surface damage rather than a weld. Document for each forge weld: steel types involved, flux type, flux application heat, weld attempt heat, outcome, and the visual assessment of the weld line on grinding.

Grain refinement and normalization mechanics

Dynamic recrystallization during hot forging

Each hammer blow applied to austenitic steel at forging temperature deforms the FCC austenite grains. The plastic deformation introduces dislocations — line defects in the crystal lattice — and at sufficient dislocation density, the stored elastic strain energy within the deformed grain exceeds the energy of creating new grain boundaries. This drives dynamic recrystallization: new, smaller, equiaxed grains nucleate at the boundaries and within the heavily deformed regions of existing grains and grow by consuming the deformed material around them, converting the large, strained grains into a population of smaller, lower-strain grains. The process is “dynamic” because it occurs during deformation rather than in a separate annealing step.

The result is grain refinement with each hammer blow at forging temperature: the forge-work that shapes the blade also continuously refines the austenite grain structure, provided the steel is above Ac3 and in the austenite phase. This is why a well-forged blade that has been worked thoroughly at heat — many controlled hammer blows distributing the deformation evenly across the section — starts with a finer grain structure than a blade that was forged with fewer, heavier blows concentrated in certain areas. The grain refinement is real and measurable: finer prior austenite grain size at quench produces finer martensite packets in the final hardened blade, and finer martensite packets mean more grain boundary area per unit volume, which inhibits crack propagation and improves toughness.

Why working below critical temperature creates problems

Below Ac1 (723°C / 1333°F), iron is no longer in the FCC austenite phase — it has reverted to BCC ferrite plus iron carbide (pearlite). Hammer blows below Ac1 do not cause dynamic recrystallization because the BCC ferrite phase does not undergo the same thermally activated grain boundary migration that drives recrystallization in austenite at hot-forging temperatures. Instead, the hammer blows create dislocations that accumulate in the ferrite without being consumed by recrystallization — this is work hardening. Work-hardened ferritic steel becomes progressively stiffer and less responsive to the hammer, requires more force to deform, and eventually cracks if continued.

The practical diagnostic for working below critical temperature: the steel begins to feel “springy” under the hammer — less plastic, bouncing slightly rather than flowing — and the surface may show hairline cracks developing. The treatment is to return the steel to full austenite temperature (above Ac3) and allow a normalizing heat to dissolve the dislocation structure, after which forging can resume normally. Work hardening in ferritic steel cannot be refined by continued forging at that temperature; it requires a normalizing heat.

Normalization procedure and what it achieves at the microstructural level

Normalization is the process of heating a blade above Ac3 (the temperature for complete conversion to austenite) and allowing it to air cool rather than quenching. For most high-carbon blade steels, the normalization temperature is Ac3 plus approximately 25–55°C (45–100°F), typically 845–870°C (1553–1598°F) for 1084 and 830–860°C (1526–1580°F) for 1095. The hold at temperature is brief — just long enough for the entire cross-section to reach the target temperature, not long enough for significant grain growth.

Air cooling from austenite temperature produces fine pearlite — a finer lamellar spacing than slow furnace cooling would produce, because the faster cooling rate (still much slower than a quench, but faster than furnace cooling) produces a finer interlamellar pearlite spacing. This fine pearlite is chemically uniform and mechanically consistent: carbides are in small, well-distributed lamellar form rather than the larger carbide clusters that can segregate to grain boundaries during forging. The normalization cycle provides several benefits: it dissolves carbide clusters that formed at grain boundaries during the thermal history of forging; it redistributes carbon more uniformly through the austenite before the air cool; and it produces a fresh, controlled grain structure from a well-defined starting state rather than from the complex strain history of forging.

Two to three normalization cycles are standard practice in blade making. The rationale for multiple cycles: each cycle dissolves any remaining carbide clusters that the previous cycle may have only partially dissolved, redistributes carbon progressively more uniformly, and produces a fresh austenite grain structure from the fine pearlite left by the previous cycle’s air cool. The final cycle produces the finest, most homogeneous pearlite starting structure for the subsequent hardening heat. The final normalization before the hardening quench is the most important: the grain size at the moment of quench determines the prior austenite grain size (PAGS), which determines the martensite packet size in the final hardened microstructure, which determines the toughness of the blade. Finer PAGS at quench produces finer martensite packets, more grain boundary area per unit volume, and a blade that resists crack initiation and propagation under lateral stress. A blade normalized once with a coarse starting grain structure before the hardening quench will be harder to toughen through tempering than a blade normalized three times with a uniformly fine grain structure at quench.

Apple Tax for blacksmithing Patreon creators

Blacksmithing and bladesmithing creator iOS rates are moderate compared to craft categories consumed primarily in couch-scroll contexts, because forge work documentation is frequently watched near the shop — on a monitor or tablet close to the anvil, or on a laptop during the post-forge review session. The iOS rate for this audience reflects a mixed consumption context: some patrons subscribe on a desktop after watching forge content on YouTube; others subscribe through iOS after watching short-form forge content on TikTok or Instagram.

Blacksmithing YouTube: 50–65% iOS. Bladesmithing YouTube with collector and knife-enthusiast audience overlap: 50–62% iOS (the collector research context draws desktop engagement). Instagram blacksmithing and bladesmithing: 70–80% iOS. TikTok forge process content: 72–82% iOS.

The Apple Tax on November 1, 2026: Patreon applies Apple’s 30% IAP fee to all subscriptions processed through the iOS Patreon app.

Even at below-average iOS rates for craft content, the dollar amounts are material. A blacksmithing creator with a YouTube-primary audience at 58% iOS loses over $835/year beginning November 1, 2026. A bladesmithing creator with an Instagram overlap at 62% iOS loses over $1,339/year. These are real dollar losses on revenue that was already earned — the same patrons paying the same subscription prices, with a different share of each payment arriving in the creator’s account.

The fix: enable Patreon’s web-only billing toggle in Creator Settings before October 31, 2026. Update all YouTube description links, the channel About page, and any bio links on Instagram or TikTok to direct to the Patreon web URL rather than the iOS app. Patrons who subscribe through a browser’s web checkout do not generate iOS-billed subscriptions and are unaffected by the Apple IAP fee. From Safari on iPhone, verify that clicking your Patreon link takes the patron to a web payment dialog, not an Apple IAP dialog, before November 1.

Use the KeepTier Apple Tax Calculator to run your specific MRR and iOS rate. KeepTier is a web-only membership page for creators who want 100% of their tier revenue minus only Stripe fees, with no iOS IAP pathway and no platform percentage. Plans from $9/month.

Patreon for blacksmithing creators — SEO guide (tiers, heat documentation, steel selection, Apple Tax table) · Patreon for woodworking creators · How the Apple Tax works · All explainers

Frequently asked questions

Related: Patreon for blacksmithing creators (SEO guide) · Patreon for woodworking creators · How the Apple Tax works · All explainers