Question 2 of 8: Why carbon embrittles high-tensile steel, and how maraging steels avoid it

Question 2: Why carbon embrittles high-tensile steel, and how maraging steels avoid it (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Given. Two routes to a high-tensile steel: the conventional one, in which strength comes from carbon dissolved in martensite and subsequently tempered, and the maraging route, in which carbon is reduced to 0.03 per cent or less and strength comes from a later precipitation treatment using nickel, cobalt and molybdenum.

Note on the stem. This paper does not name the steels it is describing: it gives only the recipe — carbon eliminated or reduced to a trace, hardening carried instead by nickel, cobalt and molybdenum. Identifying that family as the maraging steels is the first mark of the answer, and everything below is written about them.

Find. The metallurgical reasons why carbon produces brittleness, distortion, poor machinability and poor weldability in conventional steel, and the mechanisms by which the maraging composition and heat-treatment sequence remove each of those difficulties.

Part one: why carbon does the damage. Carbon strengthens steel by being trapped where it does not fit. In austenite, carbon dissolves interstitially in the relatively open octahedral sites of the face-centred cubic lattice. Quench that austenite fast enough and it transforms by a diffusionless shear to martensite before the carbon can diffuse out, so the carbon is inherited into a body-centred structure whose octahedral sites are much smaller and, crucially, are not equivalent in the three directions. Occupying only one of the three site families stretches the cell along that axis, and the lattice becomes body-centred tetragonal with a c/a ratio that grows linearly with carbon content. The resulting tetragonal strain field is enormous — a fraction of a per cent of carbon can raise the hardness from 150 to over 700 HV — and it is what pins dislocations so effectively.

That same strain field is the source of every problem the question lists. Dislocations that cannot move cannot blunt a crack tip by plastic flow, so the fracture mode changes from ductile microvoid coalescence to cleavage or intergranular separation: hardness bought this way is paid for in toughness, and the ductile-to-brittle transition temperature of a high-carbon martensite can sit well above room temperature. The transformation is also accompanied by a volume increase of roughly 4 per cent, because martensite is less dense than the austenite it comes from. Since a real part cools from the outside inwards, different regions transform at different times, and the resulting mismatch of transformation strains, superimposed on ordinary thermal contraction, produces distortion and locks in tensile residual stresses that can be large enough to crack the part during or shortly after the quench. Retained austenite in the higher-carbon grades adds a further, delayed dimensional change.

The consequences for manufacture follow directly. Machining a hardened carbon steel is difficult because it is hard and because the tempered carbides are abrasive, so the practical route is to machine soft and heat treat afterwards — but the heat treatment then moves the part, and the finished dimensions must be recovered by grinding a workpiece that is now 55 to 60 HRC. Cold forming is impracticable for the same reason in reverse: the low ductility of hardened martensite leaves no forming reserve. Welding is the worst case of all. The heat-affected zone is re-austenitised by the arc and then self-quenched by the surrounding cold mass, so a brittle, untempered, high-carbon martensite forms exactly where the joint is most restrained. Any hydrogen picked up from moisture, rust or the consumable diffuses to that hard zone and causes delayed cold cracking. Managing this needs preheat, controlled interpass temperature, low-hydrogen consumables and a post-weld heat treatment — the very expense the question refers to — and the carbon-equivalent formula $CE = C + \dfrac{Mn}{6} + \dfrac{Cr+Mo+V}{5} + \dfrac{Ni+Cu}{15}$ exists precisely to quantify how much of it is required.

Part two: what the maraging composition changes. Maraging steels remove the cause rather than manage the symptom. With carbon at or below 0.03 per cent the alloy is essentially a binary Fe–18Ni base with cobalt, molybdenum and a little titanium. Nickel is a strong austenite stabiliser, so on cooling from the annealing temperature of about 820 °C the alloy still transforms martensitically — the martensite-start temperature sits near 200 °C and the finish well above room temperature — but the product is a carbon-free, body-centred cubic lath martensite. Without interstitial carbon there is no tetragonality and no vast strain field, so the as-transformed material is soft (about 30 HRC), tough, and readily machined, formed and welded. Two further practical consequences fall out of the same fact: the transformation is so sluggish in composition terms that air cooling suffices in any realistic section, which removes the quench and with it the quench distortion and quench cracking; and because the structure is a dislocated lath martensite rather than a plate martensite, it work-hardens normally instead of cracking.

Strength is then restored, after all the shaping and joining is finished, by ageing for three to six hours at about 480 °C. During that hold the substitutional solutes precipitate as fine, semi-coherent intermetallics — Ni₃Mo and Ni₃Ti, with Fe₂Mo forming on longer ageing — on the dense dislocation network the martensite already provides. The roles of the three named elements are distinct and worth stating separately. Nickel makes the martensite possible and keeps it tough, and supplies the Ni₃X stoichiometry of the strengthening particles. Molybdenum is the principal precipitate former; it is the element that actually carries the strength. Cobalt forms no precipitate of its own — it works indirectly, by lowering the solubility of molybdenum in the martensite matrix and thereby raising the volume fraction of Mo-rich particles that can form. This synergy is why the classic grades carry 8 to 12 per cent cobalt alongside 3 to 5 per cent molybdenum, and why removing the cobalt costs far more strength than its own precipitation contribution would suggest.

The pay-off is a material with 1400 to 2400 MPa yield strength at a fracture toughness of roughly 100 MPa·m^0.5 — two to three times the toughness of a quenched-and-tempered steel of the same strength. Because substitutional atoms produce a nearly symmetric, and much smaller, distortion of the lattice than an interstitial does, and because the particles are small, numerous and coherent enough to be sheared or bypassed without generating large stress concentrations, dislocations are impeded without the crack-tip embrittlement that interstitial carbon causes. And because the ageing treatment involves no phase change of the matrix, only a precipitation reaction, the dimensional change is a shrinkage of about 0.04 to 0.10 per cent — small, uniform and predictable, so a finish-machined part can be aged and remain within tolerance. Welding is transformed for the same reason: the heat-affected zone reverts to the same soft nickel martensite, there is no hard zone and no hydrogen-cracking mechanism, no preheat is needed, and a single post-weld age at 480 °C brings weld metal, heat-affected zone and parent metal back to full strength together.

The trade-offs are cost and corrosion. Eighteen per cent nickel, nine per cent cobalt and five per cent molybdenum make maraging steel several times the price of a 4340, and the alloy has no more corrosion resistance than plain carbon steel, so it is used where the combination of very high strength, toughness, dimensional stability and fabricability justifies the expense — rocket motor casings, aircraft undercarriage components, die-casting dies, and the tooling that has to be machined complex and hardened afterwards without moving.