P92 Ingot Cracking: Why Pouring Parameters Beat Chemistry at the Mold

P92 Casts Differently Because It Transforms Differently

CSEF grade P92 is the 9% chromium, tungsten-bearing martensitic steel behind ultra-supercritical power-plant piping. Under ASME and ASTM it is SA-335 P92 (seamless pipe), with the European tube designation X10CrWMoVNb9-2 (EN 10216-2, material 1.4905) and the Chinese equivalent 10Cr9MoW2VNbBN under GB 5310-2017. The same chemistry also appears as F92 forgings and WP92 fittings. What unites every form is a tempered-martensite matrix strengthened by V-Nb-B microalloying and W-rich Laves phase — and that martensite is exactly why the steel misbehaves in the casting pit.

Unlike ferritic or pearlitic grades that cool quietly, P92 undergoes a martensitic transformation during solidification that expands the local volume. In a large ingot that expansion is not free — it is trapped inside a stiffening shell and converted into internal stress. This single physical fact, more than any impurity, is what makes big P92 ingots crack where ferritic grades would not. The latest numerical work on a 19-ton P92 ingot confirms it, and points the fix squarely at the pouring parameters.

P92 Across the Standard Map

Because the same melt is sold under different names by code, buyers should know which designation maps to which market. The chemistry is identical; only the product form and acceptance clause change.

The nominal composition that defines the family — and that the ingot study held constant — is a 9%Cr, 1.5-2.0%W, low-C (0.07-0.13%), B-N-V microalloyed balance.

Why Large P92 Ingots Crack Where They Do

The 19-ton study built a coupled temperature-stress FEM model in ProCAST on a 2720 mm ingot, simulated as a quarter for efficiency. As solidification advanced, high stress migrated from the ingot tail upward into the body. The decisive finding: stress grew from the surface inward, and it peaked not in the bulk but at the solidification cracking hot-spot — the concave point on the tail edge.

That concave point is a V-shaped geometric notch. Its positive curvature magnifies local tensile stress, while the surrounding shell and the still-liquid core impose two-directional shrinkage constraint. On top of that, P92’s martensitic transformation adds a volume-expansion push. The three effects stack, and the model clocked the concave point at roughly 425 MPa — above the material’s high-temperature critical strength, which is why it cracks there. A comparison path to a convex edge stayed safely inside the limit. Crucially, the simulated crack location matched the actual nozzle-end-face cracks seen on the shop floor, validating the model.

The takeaway is uncomfortable for anyone who treats cracking as a chemistry problem: the failure is geometric and thermal, amplified by a transformation unique to this grade. You cannot alloy your way out of a notch and a pouring window.

What the Casting Boundary Conditions Reveal

The model’s heat-transfer coefficients expose where the ingot loses heat — and therefore where gradients form. The spread is enormous, from near-adiabatic insulation at the hot top to an aggressive 800 W/m²·K at the mold bottom.

A 2600-fold difference between the insulated top and the chilled bottom is what drives the steep through-thickness gradient that feeds thermal stress. The pouring parameters are the only practical knobs a melter has to flatten that gradient before it becomes a crack.

Two Pouring Levers That Decide Cracking

The study varied pouring temperature (1540 / 1550 / 1560 °C) and pouring speed (2 / 3 / 4 t/min) and tracked stress at the high-risk concave point. Both levers moved the outcome — in the same direction.

Higher superheat prolongs solidification, steepens the inner-to-outer gradient, and raises residual stress. The temperature effect on the concave point was modest (about 4.7% stress change per step), but it compounded with speed.

Slow pouring cut the extreme stress about 9% by damping the thermal-shock effect and evening out the temperature field. The paper’s recommendation is unambiguous: pour at 1540 ± 5 °C and at low speed, trading a little cycle time for a crack-free ingot.

What This Means for Buyers of P92 Product

A cracked ingot does not stay in the casting shop. Surface and subsurface cracks from the tail-edge notch are rolled or forged outward into the final tube, forging, or fitting, where they become the starting points for in-service failure. So the pouring discipline of your P92 supplier is a property of the material you receive, not a factory detail you can ignore.

When you specify P92, X10CrWMoVNb9-2, or 10Cr9MoW2VNbBN, ask for the ingot history, not just the heat certificate:

• Confirm low-superheat, low-speed pouring is the shop standard for CSEF grades — the study shows it removes roughly 9% of the worst-case stress and lowers the hot-crack risk materially.
• Require ingot surface inspection and, where the product is critical, UT of the as-cast body before forging, because the tail-edge concavity is the predicted crack site.
• Do not treat P92 like a conventional Cr-Mo grade; its martensitic expansion demands the controlled cooling andPWHT that the grade’s own standard assumes.

The chemistry of P92 is solved science. The cracking is process science — and the process is the part a buyer can still influence by choosing a supplier who controls the pour.