UNS G41400 (aircraft quality E4140 = G41406) · W.Nr. 1.7225 · EN 42CrMo4 (free-machining variant 42CrMoS4 = 1.7227) · ASTM A29 / SAE J404 band: C 0.38-0.43% – Mn 0.75-1.00% – Si 0.15-0.35% – Cr 0.80-1.10% – Mo 0.15-0.25% – P 0.035% max – S 0.040% max – balance Fe. NICKEL IS NOT A SPECIFIED ELEMENT; in 4140 nickel is a residual. The EN 10083-3 42CrMo4 band is NOT THE SAME: C 0.38-0.45% – Si 0.40% max – Mn 0.60-0.90% – Cr 0.90-1.20% – Mo 0.15-0.30% – P 0.025% max – S 0.035% max. The two bands do not fully overlap on chromium (ASTM floor 0.80%, EN floor 0.90%; ASTM ceiling 1.10%, EN ceiling 1.20%) and are offset on manganese. Material certified to 42CrMo4 is therefore not automatically acceptable against a 4140 order; acceptance depends on the heat analysis meeting both bands. THIS IS A MARTENSITIC QUENCHED-AND-TEMPERED STEEL: it transforms to martensite on austenitising and oil quenching, and is then TEMPERED. IT IS NOT STAINLESS. It does NOT precipitation harden; there is NO H900 / H1025 / H1150 type ageing step.
Bought for machine and structural parts that are hardened by heat treatment and need medium-to-high strength together with toughness: shafts and spindles, gears, pins and bushings, hydraulic cylinder rods, die holders and plates, drilling and earth-moving components, and secondary structure and…
Forms
Round bar, flat bar, plate, sheet, pipe/tube, forging. All forms are supplied to order.
Standards
AMS (verified, plain 4140 chemistry 0.95Cr – 0.20Mo – C 0.38-0.43%): 6382 (bars, forgings, rings and stock for forging or flash welded rings; AIRCRAFT QUALITY; ANNEALED) · 6349 (BARS only; NORMALIZED) · 6529 (BARS only; SPECIAL AIRCRAFT-QUALITY CLEANLINESS; NORMALIZED) · 6381 (MECHANICAL TUBING; first issued 1942, current revision K/2022) · 6395 (SHEET, STRIP, PLATE) · 6452 (WELDING WIRE; vacuum melted; environment-controlled packaging) · 6390 (mechanical tubing – listed in the SAE catalogue with 4140 chemistry, but its current status could not be verified; see the standards note). ASTM: A29 / A29M (general requirements for hot-wrought alloy steel bars) · A322 (alloy steel bars, standard grades) · A331 (cold-finished bars) · A519 (seamless mechanical tubing) · A513 (ERW mechanical tubing) · A506 (hot-rolled sheet and strip) · A711 (stock for forgings) · A646 (premium quality blooms and billets for aerospace forgings) · A752 (wire rod) · A829 (alloy structural steel plate) · A193 B7 and B7M with A194 2H / 2HM (bolting and nuts) · A320 L7 and L7M (low-temperature bolting). EN / ISO: EN 10083-3 42CrMo4 (1.7225) and 42CrMoS4 (1.7227) · ISO 683-2. SAE: J404 · J412 · J1397. THE AMS NUMBERS ARE NOT INTERCHANGEABLE. All six carry the same chemical band (0.95Cr – 0.20Mo – C 0.38-0.43%), but they differ in PRODUCT FORM, DELIVERY CONDITION and CLEANLINESS LEVEL: 6382 is ANNEALED bar, forging and ring material and carries the…
Advantage
Its single most important practical advantage is that the same material is the base of both a structural quench-and-temper table and a bolting specification, and both guarantee numbers.
Welding
IT IS WELDABLE, BUT PREHEAT IS MANDATORY, and the delivery condition of the material governs the welding decision. PREHEAT: citing the Procedure Handbook of Arc Welding, Welding Answers gives 205-260 °C (400-500 °F) up to 13 mm (½ in) and 316-371 °C (600-700 °F) for 13-50 mm (½-2 in);
Limits
1) IT IS NOT STAINLESS. Chromium is 0.80-1.10% and no passive layer forms. Without oil, paint, phosphating, plating or another protective measure it rusts in damp air; it is not suitable for marine or chloride-bearing environments.
2) CHOICE OF TEMPERING BAND: the 250-450 °C band is not used for structural quench and temper.
Compiled from manufacturer data sheets · confirm against the current specification before ordering
On this page · click to jump
What AISI 4140 IsStandards by Product FormSpecification GapsChemical CompositionHeat TreatmentTEMPER EMBRITTLEMENTHardenability and Section Size EffectPhysical PropertiesWeldingMachiningCorrosion, Hydrogen and Sour ServiceFrequently Asked Questions
AISI 4140 is a low-alloy steel, alloyed with chromium and molybdenum, that can be hardened by heat treatment. One of the most widely used grades in the alloy steel family, it stands out for combining strength, toughness and wear resistance in a single material. Its UNS designation is G41400.
The approximately 1% chromium in the composition raises hardenability and wear resistance, while the 0.25% molybdenum increases strength at elevated temperature and reduces the risk of temper embrittlement. Together, the two make 4140 a steel that hardens through the depth of the section, so uniform mechanical properties can be achieved even in heavy cross-sections.
AISI 4140 is normally supplied hardened and tempered (T condition). The heat treatment temperatures are: forging 1150 °C, annealing 830-850 °C, normalising 870-900 °C, hardening 800-830 °C (oil or polymer quench), tempering 450-700 °C. Tempering in the 200-420 °C range is avoided in order to prevent temper embrittlement.
The grade is widely used in axles, gears, shafts, sprockets, pinions, studs and similar heavily loaded machine components. It can be supplied in round, flat, square and hexagonal bar form.
What AISI 4140 Is — and Why It Is Not “Almost Stainless”
AISI/SAE 4140 (UNS G41400 / W.Nr. 1.7225 / EN 42CrMo4 / JIS SCM440) is a medium-carbon, through-hardening chromium-molybdenum alloy structural steel. Its one distinguishing sentence: it is the most widely used quenched-and-tempered steel in the world because, instead of being excellent at one thing, it is good enough at everything — acceptable hardenability, acceptable toughness, acceptable machinability, and a low price.
And this has to be said at the outset: 4140 IS NOT STAINLESS. It contains about 1 % chromium. The threshold for forming a passive chromium oxide film is about 10.5 % — so 4140 carries roughly one tenth of what is required. The chromium is there for hardenability, not for corrosion. 4140 rusts like plain carbon steel and, left unpainted, unplated and unoiled, will show surface rust in the atmosphere within days.
The whole engineering of the alloy sits in two elements. Chromium (0.80–1.10 %) delays the pearlite and bainite transformations, which at a given cooling rate means martensite deeper into the section — that is hardenability. Molybdenum (0.15–0.25 %) does two jobs: it adds hardenability and it provides resistance to temper embrittlement. That second job is the subject of the most critical section on this page, and almost no distributor sheet mentions it.
Honest Position in the Family — Which Steel for Which Job
DEFENCE METAL
Versus 4130
Same Cr-Mo system, but nominal carbon 0.30 % (0.40 % in 4140). The result: 4130 reaches lower maximum hardness and lower strength, but is far easier to weld — its carbon equivalent is markedly lower. For welded tubular structures (aircraft frames, roll cages, chassis) 4130 is the right answer, not 4140. Choose 4130 for any job where post-weld heat treatment is impossible
Versus 4340
Same carbon (0.38–0.43 %) and similar Cr-Mo, but Ni 1.65–2.00 % is added (4140 contains NO nickel) and Mn is lower (0.60–0.80 % against 0.75–1.00 %). What the nickel buys is far greater hardenability: 4340 hardens much larger sections to the core and gives better toughness at a given strength. The cost is price and machinability. Rule: if the section exceeds ⌀100 mm, or if impact / fracture toughness is critical, move to 4340
Versus 8620 — A CATEGORY ERROR
At C 0.18–0.23 %, 8620 is a CARBURISING (case hardening) steel. It is not a through-hardening grade. Its job is to produce a hard case over a tough core; 4140’s job is to produce a uniform quenched-and-tempered structure throughout. “Which is harder, 4140 or 8620?” is the wrong question — the right question is surface or core. Use 8620 for gears, pins and bushings with surface wear and core toughness; 4140 for shafts, bars and forged bodies loaded through their volume
Versus 4142 · 4145
Same alloy system, only the carbon step differs: 4140 C 0.38–0.43 %, 4142 C 0.40–0.45 %, 4145 C 0.43–0.48 %. Higher carbon means higher attainable hardness and worse weldability. The ASTM A193 B7 bolting specification gives a C band of 0.38–0.48 % — so a B7 stud may actually be 4140, 4142 or 4145. You cannot know which without reading the certificate
These work in the 1900–2000 MPa band; 4140’s practical quenched-and-tempered band is 750–1200 MPa. They are not in the same league and the price gap exceeds a factor of ten. Maraging steels are also carbon-free and take their strength from precipitation hardening — no quench distortion, no temper embrittlement. If dimensional stability is critical, maraging; if cost is critical, 4140
The comparison ends the moment corrosion resistance is required. 17-4 PH is a martensitic stainless working in strength bands close to 4140. If the part is wet, outdoors or in a chemical environment, the choice between 4140 + coating and 17-4 PH should be made on coating life and hydrogen embrittlement risk — not on first cost
4140, 4140H and “4140 HT” — three different orders
4140 is a chemistry specification. 4140H means an H band: the chemistry widens slightly but the mill guarantees the Jominy hardenability band. If you will do the heat treatment yourself and you want repeatability, ask for 4140H — asking for “4140” guarantees no hardenability at all. “4140 HT” or “pre-hardened 4140” is neither a chemistry nor a standard but a delivery condition: bar quenched and tempered at the mill, typically shipped in the 28–32 HRC (≈285–321 HB) band. Do not put all three on one purchase line.
AMS 6395 (sheet, strip and plate) · ASTM A506 (hot-rolled sheet and strip) · ASTM A829 (alloy structural steel plate)
Forgings · rings · forging stock
AMS 6382 (forgings, rings and stock for forging or flash welded rings; annealed) · ASTM A711 (stock for forgings) · ASTM A646 (premium quality blooms and billets for aerospace forgings)
Wire · wire rod
ASTM A752 (alloy steel wire rod). No AMS WIRE number (other than welding wire) verified across four sources was found for 4140.
Welding filler metal
AMS 6452 (welding wire, 0.95Cr – 0.20Mo, C 0.38-0.43%, vacuum melted, environment-controlled packaging) · AWS A5.28 ER80S-D2 (general fabrication). AMS 6457 DOES NOT BELONG TO THIS MATERIAL: it is 4130 welding wire (C 0.28-0.33%).
Bolts · studs · nuts
ASTM A193 Grade B7 and B7M (studs and bolts) · ASTM A194 Grade 2H and 2HM (nuts) · ASTM A320 Grade L7 and L7M (low-temperature bolting, 593 °C minimum temper plus a Charpy requirement)
Sour service
NACE MR0175 / ISO 15156-2 – carbon and low alloy steels are acceptable without further testing when hardness does not exceed 22 HRC; the higher steps (26 and 30 HRC for tubulars) require SSC testing and the quenched-and-tempered condition. On the bolting side the equivalent is A193 B7M / A194 2HM.
Post-plating hydrogen embrittlement relief
SAE AMS 2759/9 (Hydrogen Embrittlement Relief / Baking of Steel Parts) – typically 190-218 °C for 2 to 24 hours, started within 1 to 4 hours of exposure.
Welding procedure group
NO ASME SECTION IX P-NUMBER IS STATED: 4140 is not listed as an ASME BPVC base metal and no P-Number could be verified across four independent sources.
In every row the AMS numbers come FIRST and ASTM and EN follow; the customer asked for the AMS numbers to be prominent. Where a product form has more than one AMS number, those are NOT alternatives: the delivery condition (annealed / normalized) and the cleanliness level differ. The AMS 6390 row is deliberately left with its caution: the number belongs to 4140 chemistry, but its current revision could not be verified. For wire (other than welding wire) and for pipe, no AMS number specific to 4140 and verified across four sources was found; that gap is stated openly.
Standards by Product Form · AISI 4140 (G41400 / 1.7225 / 42CrMo4)
DEFENCE METAL
General requirements
ASTM A29 / A29M — general requirements for hot-wrought steel bars. This is the companion document for 4140
Hot-rolled sheet and strip A506 · forging stock A711 · aerospace blooms and billets A646 (premium quality) · wire rod A752 · structural plate A829
Bolts · studs
ASTM A193 Grade B7 (standard) and B7M (sour service, ≤22 HRC). Note the B7 chemistry is a C 0.38–0.48 % band — so it may be 4140, 4142 or 4145
Nuts
ASTM A194 Grade 7 / 7M. [conflict] One publisher gives the matching nut for B7 studs as A194 Grade 2H (248–352 HB). Both are used in practice; write which nut you want on the order rather than assuming
Low-temperature bolting
ASTM A320 Grade L7, L7M, L7D — the same material with added Charpy impact requirements
AMS 6349, 6381, 6382, 6390, 6395, 6529. These numbers cover different product forms and different quality / cleanliness levels (bar, tubing, forging, premium melt) — they are not interchangeable
Military
MIL-S-5626 · MIL-S-16974 · MIL-S-46059
Europe — Q&T steel
EN 10083-3, grade 42CrMo4 (1.7225) and its sulphur-bearing variant 42CrMoS4 (1.7227). The international equivalent is ISO 683-2
Europe — bolting
EN 10269 — steels for fasteners with specified elevated and/or low temperature properties; 42CrMo4 appears there together with elevated-temperature proof strength values
NACE
MR0175 / ISO 15156-2 — for carbon and low alloy steels, a maximum of 22 HRC and a mandatory quenched-and-tempered condition. This is the single most restrictive rule on this page
Specification Gaps — What “4140” Does Not Buy You
In 4140 the gap is not in the chemistry, it is in the HEAT TREATMENT CONDITION. This is the alloy’s most trouble-prone commercial property: you buy the chemistry, but what determines the part’s properties is not the chemistry — it is the heat treatment, and the heat treatment is usually missing from the order line.
Gaps and Traps
DEFENCE METAL
“4140” alone is NOT a mechanical property commitment
ASTM A29 and A322 give chemistry and general requirements; they commit to no Rp0.2 or Rm. The same 4140 bar may, depending on delivery condition, sit anywhere between 655 MPa (annealed) and 2000 MPa (quenched, low tempered). If your order line does not state the heat treatment condition and a target hardness, you have not specified anything
The European route CLOSES this gap
EN 10083-3 defines the +QT (quenched and tempered) delivery condition with mechanical minimums per ruling section — and, beyond that, it also sets an UPPER LIMIT on Rm. Nothing of the kind exists on the ASTM side. If you want repeatability, the European ordering language is the safer one
Cast equivalent
4140 has no standardised cast counterpart. Casting specifications such as ASTM A148 and A487 are written by mechanical class, not by chemistry. The honest answer to a customer asking for “cast 4140” is: choose an equivalent CASTING CLASS and state in writing that it is not 4140
Weld metal equivalent
There is no common filler wire that reproduces 4140 chemistry, and that is deliberate. In practice ER80S-D2 is used (and undermatching ER70S-2 for toughness). Matching chemistry only makes sense on parts that will be fully re-quenched and tempered after welding. The reason is in the Welding section
Chemical Composition
SAE 4140 / UNS G41400 (ASTM A29 · A322 · SAE J404), weight %:C 0.38–0.43 · Mn 0.75–1.00 · Si 0.15–0.35 · P ≤0.035 · S ≤0.040 · Cr 0.80–1.10 · Mo 0.15–0.25 · Fe balance.
EN 42CrMo4 / 1.7225 (EN 10083-3 · ISO 683-2), weight %:C 0.38–0.45 · Si 0.10–0.40 · Mn 0.60–0.90 · P ≤0.025 · S ≤0.035 · Cr 0.90–1.20 · Mo 0.15–0.30 · Cu ≤0.40 · Fe balance. 42CrMoS4 / 1.7227: identical, with the single difference S 0.020–0.040 (with a MINIMUM).
SAE 4140 versus EN 42CrMo4 — NOT “the Same Steel”
DEFENCE METAL
Manganese — the biggest divergence
SAE 0.75–1.00 · EN 0.60–0.90. The common band is only 0.75–0.90. A heat at Mn = 0.95 % that conforms perfectly to SAE 4140 DOES NOT conform to EN 42CrMo4; a heat at Mn = 0.65 % does not conform to SAE 4140
Chromium
SAE 0.80–1.10 · EN 0.90–1.20. Common band 0.90–1.10. Again, non-conformance is possible in both directions
Carbon
SAE 0.38–0.43 · EN 0.38–0.45. EN is wider: a European heat at C = 0.44 % does not meet SAE 4140 (but does meet SAE 4142)
Molybdenum
SAE 0.15–0.25 · EN 0.15–0.30. EN is wider
Phosphorus and sulphur
SAE P ≤0.035 / S ≤0.040 · EN P ≤0.025 / S ≤0.035. EN demands a cleaner steel. This matters directly in the temper embrittlement section — phosphorus is the embrittling element
Copper
The SAE specification sets no Cu limit; EN sets ≤0.40
Practical consequence
Heats that satisfy BOTH specifications exist and are common — many mills deliberately produce dual-certified material. But that is a choice, not a rule. A page that writes “4140 = 42CrMo4” is describing the intersection of the two specifications and presenting it as an identity. If the customer asks for 42CrMo4, check the certificate against the 42CrMo4 bands LINE BY LINE
And the A193 B7 band is wider still
B7 allows C 0.38–0.48 % — i.e. it covers 4140, 4142 and 4145 together. Saying “a B7 stud is a 4140 stud” is wrong; B7 is a performance class, not a chemistry grade
42CrMoS4 — the sulphur-bearing variant and its hidden cost
42CrMoS4 (1.7227) is the same steel as 42CrMo4; its only difference is that sulphur carries a MINIMUM (0.020–0.040 %). Sulphur forms manganese sulphide (MnS) inclusions, which break the chip, lubricate the tool and improve machinability markedly. For a shop running automatic lathes, that is a real gain. Here is the cost. MnS inclusions elongate in the rolling direction and make the material anisotropic. Transverse ductility and impact toughness drop. Longitudinal Charpy may look unchanged while the transverse value has fallen significantly. So do not use 42CrMoS4 for: transversely loaded forgings, shafts under multiaxial stress, impact-loaded connections, and any part that will run at low temperature. The equation “better machinability, therefore better steel” is false for this grade, and distributor sheets routinely offer the two interchangeably.
Heat Treatment — the Heart of This Page
HEAT TREATMENT — SCHEMATIC
DEFENCE METAL
1 · HOT WORKING / FORGING
Step
1 · HOT WORKING / FORGING
Summary
Not a heat treatment but a precondition: the forging finish temperature and the cooling that follows decide whether a normalize is needed.
Temperature
The sources diverge at the upper limit. Saarstahl and Lucefin 1100-850 °C · Rodacciai 850-1150 °C · Flame Hardening 850-1050 °C. AZoM (ASM-derived) gives 926-1205 °C, which sits above the others; NO AVERAGE HAS BEEN TAKEN.
Time
Until the whole section is at temperature. No numerical time was confirmed across four independent sources, so none is stated.
Cooling
Do not forge below 850 °C. Flame Hardening says furnace cool; Saarstahl and Lucefin say cool in still air.
Resulting hardness
As-forged hardness depends on section and cooling rate; no binding hardness is stated for this step.
DEFENCE METAL
2 · NORMALIZING
Step
2 · NORMALIZING
Summary
Refines and homogenises the grain structure after forging or rolling. Recommended before hardening.
Temperature
840-900 °C. Saarstahl 850-880 °C · Doerrenberg 850-880 °C · Lucefin 870 °C · Flame Hardening 840-880 °C · Vulcan 870-900 °C · Nifty Alloys 870-900 °C. DIVERGING SOURCE: AZoM gives a single value of 913 °C (1675 °F).
Time
Vulcan gives 10-15 minutes per 25 mm of section. No single numerical time was confirmed across four independent sources, so no binding time is stated.
Cooling
AIR cool (Saarstahl, Doerrenberg, Lucefin, Flame Hardening, Vulcan and Nifty Alloys all agree).
Resulting hardness
Normalized hardness depends on section; no single value was confirmed across four independent sources, so none is stated.
DEFENCE METAL
3 · SOFT (SPHEROIDISING) ANNEAL AND FULL ANNEAL – THESE ARE NOT THE SAME THING
Step
3 · SOFT (SPHEROIDISING) ANNEAL AND FULL ANNEAL – THESE ARE NOT THE SAME THING
Summary
The soft anneal stays BELOW the critical temperature and is for machinability. The full anneal goes ABOVE it.
Temperature
SOFT ANNEAL (subcritical): 680-720 °C. Saarstahl 680-720 °C · Doerrenberg 680-720 °C · Ovako 680-720 °C · Flame Hardening 680-720 °C · Lucefin 720 °C · Rodacciai 650 °C. FULL ANNEAL (supercritical): 800-872 °C. AZoM 872 °C (1600 °F) · Ellwood 871 °C (1600 °F) · Vulcan 800-850 °C · Nifty Alloys 840-870 °C · Rodacciai gives 830-860 °C for an isothermal anneal.
Time
Until the whole section is at temperature. No numerical time was confirmed across four independent sources, so none is stated.
Cooling
SLOW FURNACE COOLING. Lucefin and Ovako give 15 °C per hour down to 600 °C; Saarstahl, Doerrenberg, Vulcan and AZoM say furnace cool.
Resulting hardness
In the soft annealed (+A) condition, 241 HBW MAXIMUM (Saarstahl and Doerrenberg). Lucefin gives separate figures for the as-rolled condition: +AR 301 HB max, controlled-cooled +ARc 279 HB max – these are NOT the annealed condition and must not be confused with it.
DEFENCE METAL
4 · AUSTENITISING + QUENCH (hardening)
Step
4 · AUSTENITISING + QUENCH (hardening)
Summary
This is the step that produces the hardness. Carbon goes into solid solution and the quench turns the structure to martensite.
Temperature
820-880 °C. Saarstahl 820-860 °C · Doerrenberg 820-860 °C · Rodacciai 830-850 °C · Flame Hardening 830-860 °C · Ovako 840-880 °C · Lucefin 860 °C · AZoM and Ellwood 845 °C (1550 °F) · Nifty Alloys 845-870 °C · Industeel about 850 °C. DIVERGING SOURCE: West Yorkshire Steel 860-890 °C. NO AVERAGE HAS BEEN TAKEN.
Time
Industeel gives one hour per 25 mm (1 in) of section. No single numerical time was confirmed across four independent sources, so no binding time is stated.
Cooling
OIL. In all ten sources listed, oil is the first choice (Saarstahl, Doerrenberg, Ovako, Rodacciai, Lucefin, Flame Hardening, AZoM, Ellwood, West Yorkshire, Nifty Alloys). Saarstahl, Doerrenberg, Ovako and Rodacciai also list WATER for heavy sections; Lucefin adds a polymer solution. A water quench raises the risk of cracking.
Resulting hardness
The as-quenched (untempered) hardness differs between sources: Lucefin measures 57 HRC at a 100 °C temper (so the as-quenched value is just above that), Nifty Alloys gives 55-60 HRC, and the Doerrenberg TTT diagram shows about 48 HRC for an oil quench from 850 °C. NO SINGLE VALUE IS STATED. In this condition the material is brittle and IS NOT USED WITHOUT TEMPERING.
DEFENCE METAL
5 · TEMPERING
Step
5 · TEMPERING
Summary
MANDATORY after quenching. The temperature is chosen for the target strength and toughness.
Temperature
STRUCTURAL QUENCH-AND-TEMPER BAND (European producer consensus): 540-680 °C. Saarstahl 540-680 °C · Lucefin 540-680 °C · Ovako 540-680 °C · Flame Hardening 540-680 °C · Rodacciai 550-650 °C · Industeel 520-640 °C. WIDE US BAND (by hardness target): 205-650 °C (400-1200 °F) – AZoM, Ellwood and Nifty Alloys. For 250-450 °C see the FORBIDDEN BAND box.
Time
Flame Hardening and Vulcan: soak at heat for AT LEAST 1 hour. Industeel recommends double tempering with a full cool to room temperature after each temper.
Cooling
Air cool (Saarstahl, Lucefin, Flame Hardening). Total Materia gives ACCELERATED cooling from above 600 °C as the countermeasure against reversible temper embrittlement.
Resulting hardness
See the tempering table.
DEFENCE METAL
Tempering table
Note
The table shows the relationship between tempering temperature and hardness/strength. EVERY ROW IS GIVEN WITH ITS SOURCE. The Lucefin rows are a MEASURED curve: a Ø10 mm specimen oil quenched from 850 °C. In a heavy section the same tempering temperature gives LOWER hardness, because full martensite does not form at the centre. This table is not an ordering specification; an order must be tied to an EN 10083-3 +QT diameter row or to a specification such as ASTM A193 B7.
DEFENCE METAL
FORBIDDEN TEMPERING BAND – 250-450 °C (for structural quench and temper)
Step
FORBIDDEN TEMPERING BAND – 250-450 °C (for structural quench and temper)
What happens
Impact toughness drops. Two separate mechanisms cover this band: tempered martensite embrittlement (TME, irreversible) and temper embrittlement (TE, reversible).
As named in the source
The Industeel 4340 data sheet states plainly ‘avoid temperature between 250-450 °C’; the same producer’s 4140 sheet gives the tempering band as 520-640 °C. Herring (The Heat Treat Doctor): TME at 250-400 °C (480-750 °F), the mechanism being cementite precipitation on prior-austenite and interlath boundaries together with impurity segregation, IRREVERSIBLE; TE at 375-575 °C (705-1070 °F), the mechanism being segregation of P, Sn, As and Sb to grain boundaries, REVERSIBLE by re-tempering above 575 °C with rapid cooling. Thermal Processing Magazine: TE 375-575 °C. Total Materia: the irreversible form at 250-400 °C, the reversible form at 450-650 °C, countered by accelerated cooling from above 600 °C and by 0.2-0.3% molybdenum.
Contrary evidence
THE SOURCES ARE NOT UNANIMOUS ON THIS BAND, and the counter-evidence is recorded too: Carpenter describes 4140 as ‘a through hardening Chromium-Molybdenum medium Carbon steel which is not subject to temper embrittlement’. Flame Hardening and Vulcan state that molybdenum keeps the steel from being susceptible to temper brittleness. Herring and Total Materia also confirm that molybdenum (0.2-0.3%) reduces the effect – the molybdenum band of 4140 is 0.15-0.25%, which reduces the effect but does not remove it. PRACTICAL CONCLUSION: the band the European producers (Saarstahl, Lucefin, Ovako, Flame Hardening) give for structural quench and temper is 540-680 °C, entirely outside the disputed region. 250-450 °C is used only where high hardness is the target and there is no impact toughness requirement.
Servis uyarisi
This band is not only a HEAT TREATMENT question: long SERVICE inside 250-450 °C runs the same mechanism.
The diagram is schematic; the time axis is NOT to scale. No published TTT/CCT curve confirmed across four independent sources was used, so no curve is drawn. THIS ALLOY IS A MARTENSITIC QUENCH-AND-TEMPER STEEL: it hardens by austenitising, oil quenching and tempering. It does NOT precipitation harden; there is NO ageing step of the H900 / H1025 / H1075 / H1150 type. The five steps below were each verified separately. 4140 DOES NOT PRECIPITATION HARDEN. There is NO ‘H900 / H1025’ type step for this material; those steps belong to precipitation hardening stainless steels such as 17-4 PH. A soft anneal (680-720 °C) and a full anneal (800-872 °C) are NOT the same thing and do not give the same result. The order must state which one is required. The quench medium is oil. Some producers allow water for heavy sections, but it raises the risk of cracking; the design must account for that. Quenched but untempered 4140 is NOT used. Tempering is the mandatory final step of the cycle. The tempering table is for a Ø10 mm specimen. In a heavy section the same temperature gives lower hardness; this is not a quality problem but a consequence of the hardenability limit. Normalized and annealed delivery conditions map to different AMS numbers: AMS 6349 and 6529 are normalized, AMS 6382 is annealed. The sources diverge on the 250-450 °C band; whoever reads this card should see that the counter-evidence is recorded too. The point of agreement is the 540-680 °C band for structural quench and temper.
With 4140 you do not buy the material, you buy the heat treatment. The same chemistry can sit anywhere between 27 HRC and 57 HRC depending on the tempering temperature. The tables below are the most-used part of this page.
Transformation Temperatures and Core Heat Treatment Parameters
DEFENCE METAL
Critical points
Ac1 745 °C · Ac3 790 °C · Ms 335 °C · Mf 80 °C. Mf at 80 °C matters: if you finish the quench at room temperature the transformation is complete; 4140 generally has no significant retained austenite problem
Normalising
870 °C, still air. [CONFLICT — important] One US source gives the result as 302 HB; a European source gives ≈190 HB for the same 870 °C. The difference comes from section thickness and the real air-cooling rate — a thin specimen can partly martensite in air. Do not promise a normalised hardness; ask for the section
Soft annealing
US practice: heat to 845 °C, cool from 755 °C to 665 °C at 14 °C/h, then in air → ≈197 HB (alternative: cool rapidly to 675 °C and hold 5 h isothermally). Europe (+A):720 °C, furnace cool to 600 °C, then air → ≤241 HB
Hardening (austenitising)
845–870 °C (1550–1600 °F). European practice 860 °C. Soak: 1 hour per 25 mm of section, minimum 1 hour
Quench medium
Agitated OIL is the standard medium; polymer is also used. Do not water quench — 4140 is prone to quench cracking in water and its hardenability is already adequate for oil
As-quenched maximum hardness
54–58 HRC (fully martensitic structure; the spread follows the 0.38–0.43 % carbon band). This is NOT a delivery condition — untempered martensite is unusable
Tempering range
204–649 °C (400–1200 °F). European +T practice 540–680 °C. Minimum soak: 2 hours after the load is at temperature throughout
Stress relief
At least 28 °C (50 °F) below the tempering temperature (European +SR: 50 °C below). Removes 60–70 % of residual stress, at a cost of 2–3 HRC over a two-hour soak
Forging temperature
≈1232 °C (2250 °F). Another publisher gives the working band as 1204 → 927 °C (2200 → 1700 °F)
The tempering curve — all the numbers, and which section they came from
The table below is a complete European tempering curve. The test condition must be stated explicitly: ⌀10 mm specimen, oil quenched from 850 °C. That is a small specimen and the values are an UPPER BOUND. A ⌀100 mm bar will reach none of these rows at its core.
Tempering Temperature → Hardness and Strength (⌀10 mm, oil from 850 °C)
There are THREE conclusions to draw from this table and none of them appear on distributor sheets. 1) Yield strength PEAKS at 200 °C (1620 MPa) and then falls. As-quenched (HRC 57) the yield is 1520 MPa — i.e. lower than tempered. The reason is the relief of internal stresses in the martensite and fine carbide precipitation during low-temperature tempering. “Tempering always weakens the steel” is false. 2) Charpy energy is FLAT between 150 °C and 450 °C (26–28 J). Across a three-hundred-degree tempering range toughness does not improve at all, only strength falls. So tempering at 350–450 °C gives up 500 MPa of strength and buys NOTHING in return. Toughness only opens up above 500 °C: 31 J at 500 °C, 42 J at 550 °C, 75 J at 600 °C, 114 J at 650 °C. 3) That gives the real rule for choosing a tempering temperature. If impact or fracture toughness matters, temper at at least 550 °C, preferably 600 °C and above. If you want only hardness and wear resistance, stay in the 200–250 °C band. The intermediate 300–500 °C band is the worst of both worlds — and, as the next section shows, the embrittlement hazards sit exactly there.
The second and third tempering tables — why they do not agree
Source B (US, secondary): 204 °C 52–54 HRC · 316 °C 48–51 · 427 °C 43–46 · 482 °C 38–42 · 538 °C 34–38 HRC. Source C (secondary): 200 °C 50–52 HRC · 300 °C 45–48 · 400 °C 40–44 · 500 °C 35–38 · 600 °C 28–32 · 650 °C 25–28 HRC. The full curve above (Source A): 200 °C 54.5 · 300 °C 51 · 400 °C 47 · 500 °C 42 · 600 °C 36 · 650 °C 31 HRC. At 400 °C the three sources say 40–44, 43–46 and 47 HRC — a 7 HRC spread; at 600 °C the gap is 8 HRC. The causes are section size, soak time, actual carbon content and quench severity; none of the three is wrong, all three come from different specimens. Never make a published tempering table an acceptance criterion — the real temperature is established with your own section, in your own furnace, on a trial piece.
TEMPER EMBRITTLEMENT — the Most Critical Section on This Page
There are THREE distinct phenomena that must be named separately. Data sheets routinely confuse them, and that confusion causes real failures. All three have a different temperature band, a different mechanism and — most importantly — a different reversibility.
Band: 260–370 °C (500–700 °F). Some sources give 250–400 °C. IRREVERSIBLE. Mechanism: (a) coarse cementite films form at prior austenite grain boundaries and between laths and act as stress concentrators; (b) retained austenite becomes mechanically unstable and transforms to untempered martensite under load; (c) phosphorus and nitrogen segregate to the boundaries. Silicon and molybdenum reduce TME by retarding cementite precipitation. Oil-quenched structures (less retained austenite) are less affected than air-cooled ones
Band: 375–575 °C (707–1070 °F). REVERSIBLE: it can be undone by heating above 575 °C for MINUTES — but only if the part is then cooled rapidly. Mechanism: phosphorus, antimony, tin and arsenic diffuse to and segregate at prior austenite grain boundaries. The result: the ductile-to-brittle transition temperature (DBTT) RISES and fracture toughness FALLS — while room-temperature hardness and tensile strength barely change at all. Which is why hardness testing WILL NOT CATCH IT
3) BLUE BRITTLENESS
Occurs in a low-temperature band by a strain aging mechanism. It affects plain carbon and alloy steels; high tin or phosphorus increases susceptibility. Practical avoidance: temper above 595 °C (1100 °F)
4) A fourth that must not be confused with these
“475 °C embrittlement” concerns only steels with 15 % chromium and above (ferritic stainless steels, the ferrite phase of duplex). It has nothing to do with 4140’s 1 % chromium. If a 4140 sheet mentions 475 °C embrittlement, that sheet is copy-paste
Two formulas that predict susceptibility in advance
Susceptibility to temper embrittlement (TE) can be predicted FROM THE CHEMISTRY, and that is a real tool you can use at the ordering stage. Two established indices exist: Watanabe J factor — compositions entered in weight %; keep J < 180. Bruscato X factor — compositions entered in ppm; keep X < 20. As a broader criterion, a PE value below 2.8–3.0 is considered adequate. The harmful elements are phosphorus, antimony, tin and arsenic.Even trace amounts below 0.01 % can trigger embrittlement.Manganese and silicon MULTIPLY their effect — which is why the J factor carries an (Si+Mn) multiplier. The role of molybdenum:small molybdenum additions reduce TE susceptibility, because they form (Mo,Fe)₃P clusters that block phosphorus from reaching the grain boundary. The “Mo” in 4140 is there for exactly this job.
And Now the Big Contradiction in the Literature
DEFENCE METAL
Claim A (one producer)
4140 “is NOT subject to temper embrittlement” — a direct quotation from a producer’s alloy description
Claim B (literature)
An independent technical report exists titled “Temper Embrittlement in 4140 Seamless Tubing”. A trade association also writes that “most common low alloy steels” are affected in the 375–575 °C band and recommends using no tempering cycle below 1100 °F (593 °C)
How to resolve it
Both are partly right, and the difference is one of DEGREE. Molybdenum genuinely reduces TE susceptibility — 4140 is far less susceptible than a Mo-free Cr-Mn steel. But “reduced” and “immune” are not the same thing: 0.15–0.25 % Mo does not tie up phosphorus completely, and a heat carrying high P/Sn/Sb/As will still embrittle
And the key point
TME (260–370 °C) is an entirely SEPARATE phenomenon and molybdenum does not remove it. Even if “4140 is not subject to TE” were true, it would not remove the danger of tempering at 260–370 °C. Any advice that does not separate the two phenomena is incomplete
THE SAFE RULE
For any 4140 part that is impact-loaded, notched, welded or will run at low temperature: do not temper in the 260–575 °C band. Choose either ≤250 °C (if you want hardness only) or ≥595 °C (if you want toughness). One source gives an even narrower warning: “avoid tempering in the 232–299 °C (450–570 °F) range for impact-loaded applications”
The single most-omitted practical detail: COOLING AFTER TEMPERING
Tempering a part at 620 °C and then furnace-cooling it takes that part STRAIGHT THROUGH the temper embrittlement band. TE is reversible — it dissolves above 575 °C — but it re-forms during slow cooling between 575 and 375 °C. Phosphorus needs time to reach the grain boundary, and a heavy section cooling in a furnace gives it exactly that. Correct practice: cool from the tempering temperature fast enough to pass quickly through 575–375 °C — in oil or water for heavy sections. This one sentence explains a large share of the “the heat treatment was correct but the part still came out brittle” cases, and it appears on almost no product page.
Hardenability and Section Size Effect
4140 is not a “deep hardening” steel; it is a “deep enough hardening” steel. That distinction becomes concrete around ⌀100 mm.
Jominy End-Quench Hardenability (ISO 683-2, grain size ≥5) — HRC
DEFENCE METAL
From the quenched end 1.5 · 3 · 5 mm
53–61 · 53–61 · 52–61
7 · 9 · 11 mm
51–60 · 49–60 · 43–59
13 · 15 mm
40–59 · 37–58
20 · 25 mm
34–56 · 32–53
30 · 35 · 40 mm
31–51 · 30–48 · 30–47
45 · 50 mm
29–46 · 29–45
THE REAL MESSAGE OF THIS TABLE
Look at the band widths. At 15 mm the accepted range is 37–58 HRC — a 21 HRC spread. So two heats that both fully conform to the specification can come out 21 HRC apart at the same point after the same heat treatment. That is not an error, that is the standard
Commercial consequence
If you want repeatable heat treatment, buy “4140H”, not “4140” — H grades narrow and guarantee the Jominy band. If your production sees hardness drift between lots, this is most likely why
Section Size Effect — EN 10083-3 / ISO 683-2 +QT Minimums
DEFENCE METAL
⌀16–40 mm
Rm 1000–1200 MPa · Rp0.2 ≥750 MPa · A ≥11 % · Z ≥45 % · KV ≥35 J
⌀40–100 mm
Rm 900–1100 MPa · Rp0.2 ≥650 MPa · A ≥12 % · Z ≥50 % · KV ≥35 J
⌀100–160 mm
Rm 800–950 MPa · Rp0.2 ≥550 MPa · A ≥13 % · Z ≥50 % · KV ≥35 J
⌀160–250 mm
Rm 750–900 MPa · Rp0.2 ≥500 MPa · A ≥14 % · Z ≥55 % · KV ≥35 J
The yield falls by 33 %
750 MPa at ⌀16 mm, 500 MPa at ⌀250 mm.Same steel, same heat treatment, same standard. This is the section size effect quantified, and it is the proof that publishing a single “4140 yield strength” figure is wrong
Rm is a BAND
EN sets not only a floor but a ceiling (e.g. 1000–1200 MPa for ⌀16–40 mm). An over-hard part can clear ASTM and still fail EN. No such upper bound exists on the ASTM side
Quench depth
Oil quenching produces martensite in sections up to about ⌀100 mm (4 in). Above that the core does not fully harden and the design must account for it — or move to nickel-bearing 4340
Typical US-route values — SINGLE SOURCE, small specimen
Normalised at 870 °C: Rm 1020 MPa · Rp0.2 655 MPa · A 17.7 % · Z 46.8 % · 302 HB. Annealed at 815 °C: Rm 655 MPa · Rp0.2 425 MPa · A 25.7 % · Z 56.9 % · 197 HB. Quenched from 845 °C + tempered at 540 °C: Rm 1075 MPa · Rp0.2 986 MPa · A 15.5 % · Z 56.9 % · 311 HB. These are NOT specification minimums but typical measurements on small specimens. The 302 HB normalised figure in particular directly contradicts the European source’s ≈190 HB for the same temperature — the cause is section size and real cooling rate.
EN 10269 — Elevated-Temperature Proof Strength for Bolting (d ≤60 mm)
DEFENCE METAL
Room temperature
Rm 860–1060 MPa · Rp0.2 ≥730 MPa · A ≥14 % · KV at +20 °C ≥50 J · at −40 °C ≥40 J · at 100 °C ≥27 J
Rp0.2 versus temperature
50 °C 720 · 100 °C 702 · 150 °C 677 · 200 °C 640 · 250 °C 602 · 300 °C 562 · 350 °C 518 · 400 °C 475 · 450 °C 420 · 500 °C 375 MPa
How to read it
At 400 °C the yield is 65 % of its room-temperature value, and at 500 °C only 51 %. These are the numbers used in design calculation, and they explain why a bolt loses its preload when hot
The ASTM counterpart
For ASTM A193 B7 the commonly quoted limit is 450 °C (840 °F) continuous service; above that, A193 B16 (Cr-Mo-V) is recommended. Lower bound: no impact testing required down to −29 °C (−20 °F); colder than that, use A320 L7
ASTM A193 B7 — Mechanical Minimums by Diameter
DEFENCE METAL
≤64 mm (2½ in)
Rp0.2 ≥725 MPa (105 ksi) · Rm ≥860 MPa (125 ksi)
64–100 mm (2½–4 in)
Rp0.2 ≥655 MPa (95 ksi) · Rm ≥795 MPa (115 ksi)
100–180 mm (4–7 in)
Rp0.2 ≥515 MPa (75 ksi) · Rm ≥690 MPa (100 ksi)
Ductility
Elongation ≥16–18 % · reduction of area ≥50 % (the source gives this as a range)
Hardness
The source gives a 235–331 HB band — single-sourced. Confirm from the current edition; we publish no single ceiling figure on this page
Heat treatment requirement
Quenched in oil (or water) and tempered at a MINIMUM of 593 °C (1100 °F).That minimum tempering temperature is no accident — it is set precisely to stay above the temper embrittlement band (375–575 °C)
What B7M changes
Maximum 22 HRC, for NACE MR0175 sour service compliance. The price is lower strength: the same bolt, softer. Saying “let us substitute B7M for B7, it is the same steel anyway” means reducing the design load
Physical Properties
STRENGTH VALUES
DEFENCE METAL
Condition
Hardness
Yield MPa
Tensile MPa
Elongation
EN 10083-3 · 42CrMo4 · +QT · up to 16 mm diameter
–
900 min
1100-1300
10% min
EN 10083-3 · 42CrMo4 · +QT · 16-40 mm diameter
–
750 min
1000-1200
11% min
EN 10083-3 · 42CrMo4 · +QT · 40-100 mm diameter
–
650 min
900-1100
12% min
EN 10083-3 · 42CrMo4 · +QT · 100-160 mm diameter
–
550 min
800-950
13% min
EN 10083-3 · 42CrMo4 · +QT · 160-250 mm diameter
–
500 min
750-900
14% min
EN 10083-3 · 42CrMo4 · +A (soft annealed)
241 HBW max
–
–
–
ASTM A193 Grade B7 · bolting, up to 63.5 mm (2½ in) diameter
ASTM A193 Grade B7M · low-hardness bolting for sour service, up to 101.6 mm (4 in)
235 HBW / 99 HRB max · 100% of the batch is hardness tested
550 min
690 min
16-18% min (by size)
ASTM A320 Grade L7 · low-temperature bolting, up to 63.5 mm (2½ in) diameter
321 HBW / 35 HRC max
725 min
860 min
16% min
EVERY ROW IS A SPECIFICATION MINIMUM – a guaranteed floor, not a typical value. The akma_mpa and cekme_mpa columns carry ONLY the MPa value; the ksi equivalents are in the akma_ksi and cekme_ksi columns. The EN rows depend on DIAMETER: the same material gives a lower yield in a heavy section because hardenability is limited. The ASTM A193 / A320 rows are BOLTING specifications and include a hardness CEILING; the EN rows have no hardness ceiling. In the chart, a value given as a range is drawn at its lower bound; the full range is in the table. The EN 10083-3 rows and the ASTM A193 rows CANNOT be compared with each other: one is a structural quench-and-temper table, the other a bolting specification, and they carry different test-piece and measurement rules. B7 and B7M are made from the same material; the difference is the heat treatment target and the hardness ceiling. Sour service requires B7M. In the EN +QT rows the yield falls as the diameter grows: 900 MPa at 16 mm, 500 MPa at 250 mm. That is the numerical expression of the hardenability limit of 4140. Typical annealed tensile/yield values have NOT been put in the table: every source found for them derives from the same ASM database, so the four-independent-source requirement is not met (see the omissions list).
Physical Properties · AISI 4140 / 42CrMo4
DEFENCE METAL
Density
7.85 g/cm³ (European source) · 7.83 g/cm³ (US source, as specific gravity). The difference is negligible; use 7.85
Modulus of elasticity
≈217 GPa at room temperature, falling to 164 GPa with temperature. [conflict] A US source gives 33 Mpsi (≈228 GPa). Do not treat the modulus as constant — the drop must be accounted for in hot service calculations
Thermal expansion
European source 10.5 → 14.4 × 10⁻⁶/K (−100 to 600 °C). US source 12.2 × 10⁻⁶/K (20–100 °C, oil hardened and tempered). The two sources are consistent
Thermal conductivity
45.1 → 34.4 W/m·K (falling with temperature). US source 42.7 W/m·K at 100 °C. About three times that of stainless — which means heat escapes into the workpiece instead of burning the tool
Specific heat
423 → 587 J/kg·K (rising with temperature). US source 473 J/kg·K for 20–200 °C
Electrical resistivity
0.231 → 0.806 Ω·mm²/m, i.e. 23.1 → 80.6 µΩ·cm (strongly rising with temperature). US source 22 µΩ·cm at 20 °C
Melting point
≈1416 °C (2580 °F) — single source
MAGNETIC BEHAVIOUR
Ferromagnetic. Magnetic in every heat treatment condition. It cannot be used in any application requiring non-magnetic material — the absolute dividing line between this steel and austenitic stainless
Welding — Preheat Is MANDATORY
4140 is a difficult steel to weld, and that is the material’s single biggest practical limitation. The reason can be given in one word: hardenability. The heat-affected zone (HAZ) is austenitised and then cooled by the surrounding cold mass at quenching rates. The result is untempered, hard, brittle martensite — and combined with the hydrogen present, it produces cold (delayed) cracking.
Carbon equivalent — calculate it yourself
The measure of weldability is the IIW carbon equivalent: CE = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15 Applying that formula to 4140’s specification chemistry gives (this is a CALCULATED value, not taken from a published source): lower corner (C 0.38 · Mn 0.75 · Cr 0.80 · Mo 0.15) ≈0.70; nominal (C 0.40 · Mn 0.88 · Cr 0.95 · Mo 0.20) ≈0.78; upper corner (C 0.43 · Mn 1.00 · Cr 1.10 · Mo 0.25) ≈0.87. The common engineering threshold is CE = 0.45, above which preheat is considered necessary. 4140 sits at nearly twice that. Which is why preheat is not a recommendation but a requirement. For comparison, 4130 carries about 0.30 % carbon and its CE is markedly lower — if a welded structure is required, that is the right steel.
Welding Parameters · 4140 — THE PUBLISHED RANGES CONFLICT; WE GIVE THEM ALL
DEFENCE METAL
Preheat — four separate published ranges
Source A: ≤12.7 mm 205–260 °C, 12.7–50 mm 316–371 °C · Source A’s lower alternative: ≤12.7 mm 177 °C, 12.7–25 mm 232 °C, 25–50 mm 260 °C · Source B:288–427 °C for thick components · Source C (general):170–350 °C
Honest summary
The published range spans 175 °C to 425 °C in total. Publishing a single number would be misleading. The correct value depends on thickness, degree of restraint, hydrogen level and filler choice and is established in weld procedure qualification
HOW to apply preheat
The heat must pass through the ENTIRE section and extend at least 75 mm (3 in) in every direction from the joint. Heating only the top of the bead is not preheat
Interpass temperature
Must be held at or above the preheat temperature. Do not let it fall — every drop is a quenching cycle
Filler metal
Primary: ER80S-D2. Undermatching (for toughness): ER70S-2. If the part will be fully re-quenched and tempered after welding, a filler matching the base metal chemistry is required. WARNING: do not use a filler that OVERMATCHES the tensile strength — ductility falls and cracking risk in the weld metal rises
Hydrogen control (DHT)
Immediately after welding, hold the part at preheat temperature and wrap it in ceramic insulation: 30 minutes to 1 hour per 25 mm of thickness. This lets hydrogen diffuse out and is the single most effective measure against delayed cracking. Also use low-hydrogen electrodes and respect the baking and storage discipline
PWHT (stress relief)
566–677 °C (1050–1250 °F), ≈1 hour per 25 mm. Recommended for thicknesses above 3 mm (⅛ in)
PWHT on ALREADY heat-treated (4140 HT) parts
Must stay BELOW the original tempering temperature, otherwise the part loses hardness. [UNIT ERROR WARNING] One publication writes this as “60 °F (15 °C) below” — a 60 °F DIFFERENCE is 33 °C, not 15 °C. The source’s own conversion is wrong. In practice, stay 30–55 °C below
Cooling
Cool slowly using heating blankets. Rapid cooling means embrittlement and hydrogen-induced cracking. But after PWHT, pass quickly through 575–375 °C — remember the temper embrittlement section. These two look contradictory; they are not, they describe different temperature bands
What actually goes wrong
1. No preheat, or insufficient preheat. Single most common cause of failure in 4140. The part looks hot at the moment of welding, but the rest of the mass is cold and quenches the HAZ within seconds. The result is a delayed crack appearing days later — in transit, at assembly, or on first loading. 2. Hydrogen. Damp electrodes, oily surfaces, rust, paint residue and humid air are all hydrogen sources. Hard HAZ + hydrogen + residual stress is precisely the recipe for cold cracking. Bake the electrodes, clean the joint, apply the post-weld hold (DHT). 3. Skipping post-weld heat treatment. Without PWHT the HAZ remains untempered martensite. Test its hardness and you will read 50+ HRC — and on a NACE-scope part that is a direct non-conformance (limit 22 HRC). 4. Overmatching filler. Weld metal significantly stronger than the base metal reduces ductility and moves the crack into the weld metal. 5. Fabricating welded parts from pre-hardened bar. If you bought 28–32 HRC “4140 HT” bar, welding destroys the heat treatment condition locally. A welded 4140 part must be re-quenched and tempered as a whole after welding — or the design must be based on the weakest region.
Machining
4140 is one of the better-machining quenched-and-tempered steels — but that depends entirely on the delivery condition. Annealed (≈195 HB) it is a comfortable material; quenched and tempered (300+ HB) it is a different job altogether.
Machinability and Starting Cutting Parameters
DEFENCE METAL
Machinability rating
61 % general · 70 % in the annealed condition (against B1112 free-machining steel = 100 %) — single source
Hardness by delivery condition
Annealed ≈195 HB · Normalised 240–300 HB · Quenched and tempered 230–340 HB. The speeds below are for nominal (annealed / normalised) hardness and must be reduced substantially above 300 HB
These speeds assume stable clamping, quality material, short tool overhang and nominal hardness. If those conditions are not met, reduce the speed
The thermal advantage
Thermal conductivity is ≈43–45 W/m·K — about three times that of stainless. Heat escapes into the part instead of accumulating in the tool. This is largely the physical reason 4140 machines comfortably compared with stainless
The sulphur-bearing variant
42CrMoS4 machines noticeably better (chip breaking and tool life). The price is transverse toughness — see the chemistry section
After hardening
Finishing operations are required after hardening (grinding, hard turning). If tolerances are critical, add heat treatment distortion to your machining allowance
Corrosion, Hydrogen and Sour Service — the “NOT Stainless” Section
Atmospheric and aqueous corrosion
4140’s corrosion resistance is, in practice, that of plain carbon steel. Its 0.80–1.10 % chromium is not enough to form a passive film — the threshold is around 10.5 %. Molybdenum contributes nothing meaningful to corrosion either; both are hardenability elements. Consequence: 4140 parts must be protected — paint, zinc or zinc-nickel plating, phosphating plus oil, cadmium (in aerospace), nitriding or chrome plating. And the protection itself brings the risk described in the next section.
HYDROGEN EMBRITTLEMENT — the number one failure mode of high-strength 4140
As a steel’s strength rises, its susceptibility to hydrogen embrittlement increases SHARPLY. The practical threshold is taken as roughly 1000–1100 MPa tensile strength or 32–35 HRC — and 4140 works right at and above that band. Where the hydrogen comes from:acid pickling, electroplating (zinc, cadmium, chromium), phosphating, electrolytic cleaning, cathodic protection, and in service the corrosion reaction itself. The mechanism is insidious for this reason: the part looks normal coming out of the plating line, its hardness is correct, it passes a tensile test. Fracture arrives days later, under static load, without warning — which is why it is called delayed fracture. The countermeasure — post-plating bake-out: a long, low-temperature bake carried out immediately after plating lets trapped hydrogen diffuse out. The governing specifications are ASTM B850, ASTM F1940 and ISO 4042.The bake temperature and duration could NOT be independently verified in this research — we publish no numbers. The temperature, the duration and the maximum permissible delay after plating must be taken from the applicable specification according to the part’s strength class.What you must not do is take a random temperature/time pair from a distributor sheet and apply it.
Sulphide stress cracking and NACE MR0175
In environments containing hydrogen sulphide (H₂S), 4140 suffers sulphide stress cracking (SSC). The mechanism is a sibling of hydrogen embrittlement: the corrosion reaction generates hydrogen, the sulphide ion prevents that hydrogen from recombining into molecules and escaping, and the hydrogen enters the steel. The NACE MR0175 / ISO 15156-2 rule is unambiguous: for carbon and low alloy steels, a maximum of 22 HRC (approximately 250 HV / 237 HBW), and the material must be in the QUENCHED AND TEMPERED condition. That single rule governs the whole design of 4140 in sour service. Reaching 22 HRC means, per the tempering curve above, tempering roughly in the 620–650 °C band, which brings Rm down to around 900–1000 MPa. In other words none of 4140’s high-strength conditions can be used in sour service. And the 22 HRC limit is a LOCAL limit: weld metal, HAZ, cold-formed threads, rolled thread roots and any non-decarburised surface layer must each individually be below it. A part whose average hardness is 20 HRC can have a 45 HRC HAZ and be non-conforming. The ASTM A193 B7M and A320 L7M grades exist precisely to meet this limit.
Where not to use it
1. Wet or outdoor service without protection. It rusts. Coating, paint or oil is mandatory. 2. Sour service above 22 HRC. A NACE violation and a real cracking risk. 3. Welded structures where heat treatment is impossible. CE ≈0.78; if preheat and PWHT cannot be applied, choose 4130. 4. Impact-loaded parts tempered in the 260–575 °C band. The embrittlement bands. 5. Sections above ⌀100 mm requiring full through-hardening. The core will not harden; 4340 is needed. 6. Anywhere non-magnetic material is required. It is ferromagnetic. 7. High-strength parts electroplated without a bake-out. Delayed fracture. 8. Continuous service above 500 °C. The yield falls to half its room-temperature value at 500 °C; 4140 is not a creep steel.
Frequently Asked Questions
The customer wants material certified to 42CrMo4 and we have SAE 4140 in stock. Can we ship it?
Not without reading the certificate. The two specifications overlap but are NOT IDENTICAL, and non-conformance is possible in both directions. The most critical divergence is manganese. SAE 4140 requires 0.75–1.00 %, EN 42CrMo4 requires 0.60–0.90 %. The common band is only 0.75–0.90 %. So a heat at Mn = 0.96 % that conforms perfectly to SAE 4140 does not conform to 42CrMo4. The second divergence is chromium: SAE 0.80–1.10 %, EN 0.90–1.20 % — common band 0.90–1.10 %. A SAE heat at Cr = 0.85 % again fails 42CrMo4. It also works in reverse: the EN carbon band is 0.38–0.45 %, SAE’s is 0.38–0.43 %. A European heat at C = 0.44 % meets 42CrMo4 but fails SAE 4140 (it meets SAE 4142). And EN demands a cleaner steel: P ≤0.025 % and S ≤0.035 %, against SAE’s ≤0.035 % and ≤0.040 %. The good news is that many mills know this and deliberately produce narrow-band heats that satisfy both specifications at once — “dual certified 4140/42CrMo4” is a common product. But that is a choice, not a rule, and it does not hold for every heat. What to do: pull the mill certificate for the heat you hold and compare the Mn, Cr, C, P and S lines against the 42CrMo4 bands ONE BY ONE. If all conform, dual certification can be issued. If not, supply the customer with 42CrMo4. “It is the same steel” is not a defensible sentence in a third-party audit.
The shaft must carry 600 MPa yield and will take impact loads. What tempering temperature? The supplier suggests 400 °C.
400 °C is definitely wrong, and for two compounding reasons. The supplier’s recommendation was read off a hardness table without looking at the toughness data. First reason — toughness does not improve at all. Look at the full tempering curve: Charpy energy is 27 J at 150 °C, 26 J at 350 °C, 26 J at 400 °C, 27 J at 450 °C. Across a three-hundred-degree range toughness is effectively CONSTANT. So tempering at 400 °C drops the yield from 1620 to 1440 MPa and buys no toughness whatsoever. Toughness only opens up above 500 °C: 31 J at 500 °C, 42 J at 550 °C, 75 J at 600 °C, 114 J at 650 °C — roughly a fourfold difference. Second reason — 400 °C sits squarely inside the embrittlement band. The reversible temper embrittlement band is 375–575 °C. In that band phosphorus, tin, antimony and arsenic diffuse to and segregate at prior austenite grain boundaries; the ductile-to-brittle transition temperature rises and fracture toughness falls. And the most dangerous aspect of this damage is that it barely changes hardness or tensile strength at all — so it passes your routine quality control without a flag and reveals itself only in the field, under impact loading. The right answer: for 600 MPa yield with impact loading, temper in the 620–650 °C band. Per the curve, 650 °C gives you Rp0.2 ≈870 MPa, Rm ≈980 MPa and KV ≈114 J — well above the target yield with four times the toughness. And do one more thing: after tempering, cool rapidly through 575 → 375 °C (in oil for heavy sections). Furnace cooling re-creates on the way down the embrittlement you dissolved at 650 °C. This one sentence explains a large share of “the heat treatment was correct but the part came out brittle” cases. Remember the curve above comes from a ⌀10 mm specimen; validate on a trial piece for your own section.
Is it true that “4140 already contains chromium, so it behaves almost like stainless in mildly corrosive service”?
No, and this misunderstanding generates real costs. 4140 contains 0.80–1.10 % chromium. For a steel to behave passively — to build a self-repairing chromium oxide film — the threshold is about 10.5 % chromium. 4140 sits at one tenth of that. The chromium here is present for hardenability, not for corrosion: it delays the pearlite transformation so that an oil quench produces martensite deeper into the section. The same applies to molybdenum. In practice 4140 rusts like plain carbon steel. Protection is mandatory: paint, zinc or zinc-nickel plating, phosphate and oil, cadmium in aerospace, or nitriding. And here the real trap begins. Most of those protections are electrolytic, which means they charge the steel with hydrogen — as does acid pickling. A high-strength 4140 part (roughly above 1000–1100 MPa or above 32–35 HRC) is sharply susceptible to that hydrogen. The part leaves the plating line sound, its hardness is correct, it passes test — and days later it fractures under static load without warning. This is called delayed fracture and it is the number one field failure of high-strength 4140. So frame the decision this way: if the part is wet, outdoors or in a chemical environment, choose between “4140 + coating” and a stainless (for example 17-4 PH) not on first cost but on coating life, bake-out discipline and hydrogen risk. If you settle on 4140, make post-plating bake-out an order requirement and state which specification governs it (ASTM B850, F1940 or ISO 4042). Do not let anyone apply an arbitrary temperature and time.
We use the same heat treatment recipe but hardness drifts 6–8 HRC from lot to lot. Is the furnace faulty?
Most likely it is not the furnace — it is THE STEEL ITSELF. And it is a drift the standard permits. Look at the Jominy table. The hardenability band ISO 683-2 accepts for 42CrMo4 at 15 mm from the quenched end is 37–58 HRC. That is a 21 HRC spread, and both 37 and 58 are fully conforming. At 20 mm the band is 34–56, at 25 mm 32–53 HRC. So two conforming heats can come out more than 20 HRC apart at the same point after the same cycle in the same furnace. The 6–8 HRC you are seeing is a perfectly ordinary slice of that band. The cause is the freedom inside the chemistry: C 0.38–0.45 %, Mn 0.60–0.90 %, Cr 0.90–1.20 %, Mo 0.15–0.30 %. A heat at the top corner of those bands genuinely hardens differently from one at the bottom. Add grain size variation on top (the table assumes grain size ≥5). The fix has three steps. First and most important: buy “4140H”, not “4140”. H grades widen the chemistry slightly but narrow and GUARANTEE the Jominy band. If you do your own heat treatment, this is the product you should be buying, and most buyers do not know it. Second: run a trial piece from every lot and set the tempering temperature from it, rather than using one fixed recipe. Third: record the actual C, Mn, Cr and Mo from each mill certificate; after a few lots you will have your own hardenability correlation. And check two more things: the austenitising soak should be 1 hour per 25 mm of section, and the tempering soak at least 2 hours AFTER the load reaches temperature. The difference between “the furnace is at temperature” and “the part is at temperature” is worth several HRC on its own in heavy sections.
Common data sheet errors — check before you order
1. “4140 = 42CrMo4” — INCOMPLETE. The Mn bands overlap only between 0.75 and 0.90 %; the Cr bands only between 0.90 and 1.10 %; EN carbon reaches 0.45 % while SAE stops at 0.43 %; EN P and S limits are tighter. Dual-certified heats are common but not guaranteed. 2. A single “4140 yield strength” figure is published. Per EN 10083-3 the same steel yields 750 MPa at ⌀16 mm and 500 MPa at ⌀250 mm — a 33 % difference. Any mechanical value without a stated section is meaningless. 3. Tempering tables published without a section size. The full curve above comes from a ⌀10 mm specimen. The Rm 2200 MPa / 57 HRC row can NEVER be reached at the core of a ⌀100 mm bar. 4. “As-quenched 4140: Rm 2200 MPa” written as a product property. Untempered martensite is not a usable delivery condition — it cracks. Note also that this row carries the lowest Charpy value in the table, 24 J. 5. “4140 is not subject to temper embrittlement” — CONTRADICTED. One producer writes this; against it stands an independent technical report titled “Temper Embrittlement in 4140 Seamless Tubing”, and a trade association recommending no tempering cycle below 1100 °F (593 °C). The truth: molybdenum REDUCES susceptibility but does not confer IMMUNITY. 6. Two different embrittlement phenomena conflated.Tempered martensite embrittlement is at 260–370 °C and IRREVERSIBLE; temper embrittlement is at 375–575 °C and REVERSIBLE (dissolved above 575 °C in minutes). Different mechanism, different remedy. 7. If a 4140 sheet mentions “475 °C embrittlement”, that sheet is copy-paste. 475 °C embrittlement concerns steels with 15 % chromium and above; it has nothing to do with 4140’s 1 %. 8. “Temper at 400 °C for a good balance” — WRONG. The Charpy data is flat at 26–28 J between 150 and 450 °C. In that band you sacrifice strength and gain no toughness at all. Toughness only opens up above 500 °C. 9. Stress relief given as “60 °F (15 °C) below the original tempering temperature” — UNIT ERROR.A 60 °F DIFFERENCE is 33 °C. A published source’s own conversion is wrong. 10. Normalised hardness given as a single number. For the same 870 °C one source writes 302 HB and another ≈190 HB — the difference comes from section and real cooling rate. Likewise the modulus of elasticity is given as constant: it falls from 217 GPa to 164 GPa, and a US source writes ≈228 GPa. 11. Tempering tables diverge by 6–8 HRC between publishers, and Jominy values are given as single numbers — the standard’s band at 15 mm is 37–58 HRC. Make none of them an acceptance criterion. 12. “4140 contains chromium, therefore it resists corrosion” — WRONG. The passivity threshold is 10.5 % Cr; 4140 has 0.80–1.10 %. It rusts. 13. Preheat given as a single number. The published range spans 175–425 °C and depends on thickness, restraint, hydrogen level and filler. 14. The post-weld hydrogen hold (DHT) step is omitted. Holding the part at preheat temperature for 30 minutes to 1 hour per 25 mm after welding is the single most effective step against delayed cracking. 15. “A B7 stud is a 4140 stud” — INCOMPLETE. The A193 B7 chemistry is a C 0.38–0.48 % band covering 4140, 4142 and 4145 together. B7 is a performance class. 16. Comparison with 8620 made on the wrong axis. 8620 is a carburising steel (C 0.18–0.23 %); 4140 is a through-hardening steel. 17. The NACE 22 HRC limit read as an “average hardness”. The limit is LOCAL: weld metal, HAZ, rolled thread roots and cold-formed regions must each be below 22 HRC. 18. 42CrMoS4 offered as a drop-in for 42CrMo4. Sulphur improves machinability but reduces transverse ductility and impact toughness. 19. “4140” assumed to be the same as “4140H”.H grades guarantee the Jominy band; plain 4140 does not. If you do your own heat treatment, the difference shows up as 6–8 HRC of lot-to-lot drift.
COMPARISON
THREE SEPARATE CRITERIA, EACH READ FROM ONE SOURCE FAMILY. (1) CHEMISTRY: the band for all three grades is taken from ASTM A29 / SAE J404, so nickel, chromium and molybdenum are compared under the same specification logic. (2) DEPTH OF HARDENABILITY: for 4140 and 4340 the condition letter – section – tensile band mapping of THE SAME STANDARD (AS 1444-1996) is used; the criterion is the question ‘up to what section can the same tensile band be reached’, because that mapping holds the section and the test rule constant. (3) AMS COVERAGE: for each grade, the numbers verified from SAE title records together with their melting and cleanliness requirements. Hardness or tensile figures from DIFFERENT sources have NOT been placed side by side.
DEFENCE METAL
Grade
UNS
W.-Nr.
Carbon
Nickel
Chromium
Molybdenum
Hardenability
Toughness
Ams kapsami
Typical use
AISI 4140
G41400 (aircraft quality E4140 = G41406)
1.7225 / 42CrMo4 (close equivalent, the bands are not identical)
0.38-0.43%
NOT SPECIFIED (residual element)
0.80-1.10%
0.15-0.25%
In the AS 1444 condition table only the R/S conditions (700-930 MPa tensile) hold at a 250 mm section; the highest condition, W, is limited to 20-30 mm (1000-1230 MPa).
Izod 27-54 J in the AS 1444 table; KV 35 J minimum in EN 10083-3 +QT (16-250 mm).
6349 · 6381 · 6382 · 6395 · 6529 · 6452. There is NO VAR (vacuum arc remelted) AMS number; the highest cleanliness requirement is the ‘special aircraft-quality cleanliness’ wording of 6529.
Shafts and spindles, gears, hydraulic cylinder rods, die holders, ASTM A193 B7 studs and A320 L7 bolting, drilling components.
AISI 4340
G43400 (aircraft quality E4340 = G43406)
The W.Nr. assignment is CONTRADICTORY between sources (1.6565 / 40NiCrMo8-4 is given by some, 1.6511 / 36CrNiMo4 and 1.6582 / 34CrNiMo6 by others)
0.38-0.43%
1.65-2.00%
0.70-0.90%
0.20-0.30%
In the AS 1444 condition table condition T (850-1000 MPa tensile) holds at a 250 mm section and the range extends up to condition X (1150-1300 MPa at 30 mm). In tables where THE SAME PUBLISHER applies THE SAME STANDARD in THE SAME FORMAT, 4340 stays one condition above 4140. (The sources are not fully unanimous on this point; see the contradictions list.)
Izod 21-54 J in the AS 1444 table; nickel raises low-temperature toughness, which is why the ASTM A320 class with a -101 °C Charpy requirement, L43, is based on 4340.
6359 · 6409 · 6414 · 6415 · 6454 · 6484 (plus 6456 welding wire, 4340Mod). THERE ARE VAR-REQUIRED AMS NUMBERS (6414 and 6454) and 6409 carries a ‘special aircraft quality cleanliness’ requirement.
Aircraft landing gear, transmission and drive shafts, heavy-section crankshafts and gears, high-strength fasteners, ASTM A320 L43 low-temperature bolting.
AISI 8740
G87400
No W.Nr. equivalent verified across four sources was found
0.38-0.43%
0.40-0.70%
0.40-0.60%
0.20-0.30%
Its nickel and chromium levels sit between 4140 and 4340. No numerical section-versus-strength table could be verified across four independent sources, so NO NUMBER IS GIVEN in this row.
Izod 41 J in the annealed condition (AZoM, ASM-derived). No four-source figure was found for the quenched-and-tempered condition.
The one decisive chemical difference is NICKEL. In 4140 nickel is NOT specified; in 8740 it is 0.40-0.70%; in 4340 it is 1.65-2.00%. Nickel forms no carbide, strengthens the ferrite in solid solution and lowers the transformation temperature of austenite; the practical consequences are two: deeper hardening at the same carbon level, and higher low-temperature toughness. Carbon is 0.38-0.43% in all three grades, so the MAXIMUM ATTAINABLE HARDNESS is similar in all three. THE DIFFERENCE IS NOT IN THE HARDNESS BUT IN HOW DEEP INTO THE SECTION THAT HARDNESS REACHES.
Ams farki
The difference that decides the order in practice: 4340 has AMS numbers with a VAR (vacuum arc remelted) requirement (6414 for bars/forgings/tubing, 6454 for sheet/strip/plate) and also 6409 with a ‘special aircraft quality cleanliness’ requirement. 4140 HAS NO VAR-REQUIRED AMS NUMBER; its highest cleanliness requirement is the ‘special aircraft-quality cleanliness’ wording that AMS 6529 places on normalized bar. A specification that calls for remelted material cannot be met with 4140.
Ortak sinir
ALL THREE ARE QUENCH-AND-TEMPER STEELS AND NONE OF THEM IS STAINLESS. Corrosion protection, hydrogen embrittlement measures (the AMS 2759/9 bake after plating) and the 22 HRC limit of NACE MR0175 / ISO 15156-2 apply to all three alike. Choosing between the grades does not solve a corrosion problem.
The comparison is made WITHOUT INVENTING NUMBERS: no hardenability figure is given in the 8740 row, because no section-versus-strength table could be found across four independent sources. The hardenability difference between 4140 and 4340 is read from the section assignments that one publisher gives for the two grades in the AS 1444 condition tables. One publisher (Vulcan) gives the SAME table for both grades; that contradiction is recorded. The consequence of the nickel difference is not ‘harder’ but ‘hard deeper in’. All three grades have carbon in the same band. The difference in AMS coverage (a VAR-required number exists or does not) is the most concrete discriminator in practice.
