Longitudinal cracks in peritectic slabs: superheat, mould flux or taper?
A slab caster with a 230 mm × 1,500 mm mould is casting a 0.13% C, 1.10% Mn steel at 1.25 m/min with 38 K superheat. The mould flux in use has basicity 1.35, crystallisation temperature 1,180 °C and viscosity 0.12 Pa·s at 1,300 °C. Mould taper is 1.0%/m and broad-face water flow 4,000 L/min at a 6 K temperature rise. Longitudinal cracks are appearing 60–100 mm inboard of both narrow-face corners on the broad faces. Oscillation-mark depth has increased to 0.9 mm from a historical 0.5–0.6 mm. Sticking alarms have fired twice in a week with no breakout. The submerged entry nozzle is new, no clogging is seen, and mould level control is steady. The machine has a vertical section of 900 mm, and a mould thermocouple map exists at three levels.
1. Explain the mechanism linking this carbon content, shell contraction, the slag film and the taper to off-corner longitudinal cracks.
Read the worked answer
Step 1 - the grade. At 0.13% C the steel sits squarely in the peritectic range (roughly 0.09–0.17% C). On cooling it passes through L + δ → γ with a volumetric contraction of about 0.38% in the solid shell. That contraction occurs in the first 20–40 mm of the mould, exactly where the shell is thinnest and weakest.
Step 2 - the gap. The shell pulls away from the copper, creating an air gap. Heat transfer across the gap is dominated by radiation and by the slag film, so the local heat flux collapses, the shell stays thin and hot, and the surface reheats. The shell must then re-contact and re-shrink down the mould - the oscillation-mark depth of 0.9 mm is evidence of a thick, uneven slag film and late, lumpy contact.
Step 3 - why the flux makes it worse. A basicity of 1.35 with a 1,180 °C crystallisation temperature produces a thick crystalline slag film. Crystal layers resist heat transfer, so the shell is thinner and hotter than it should be, and friction rises. Slow heat extraction plus 38 K superheat - high for this section - keeps the shell thin at the meniscus where the crack originates.
Step 4 - why off-corner. The narrow face contracts more than the broad face relative to the fixed 1.0%/m taper, and the corner region has a two-dimensional cooling geometry. Off-corner is where the gap opens most, so the shell is thinnest and the surface temperature peaks. The longitudinal crack is a hot, tensile failure of that thinned shell just below the meniscus.
2. Rank three corrective actions - reducing superheat from 38 K to 20 K, switching to a lower-basicity/lower-crystallisation flux, and increasing taper from 1.0 to 1.4%/m - and state the side effect of each.
Read the worked answer
Rank 1 - cut superheat to 20 K. Superheat controls how much heat the shell must remove before it is strong. Every 10 K removed thickens the shell at the meniscus and narrows the liquid core, directly reducing crack susceptibility. Side effects: low superheat risks nozzle freeze and skull build-up on the stopper or SEN during ladle changes and at low casting speeds; it also needs tighter tundish temperature control (a 10 K window). No friction or lubrication penalty - which is why it goes first.
Rank 2 - increase taper to 1.4%/m. Taper is matched to the actual shrinkage of the shell; more taper holds the shell against the copper through the contraction window, keeps the gap closed and cuts the off-corner hot spot. Side effects: over-taper grips the shell. Withdrawal force and mould friction rise, which risks sticking, transverse or diagonal cracks, and in the extreme a breakout. With a 1,180 °C crystallisation flux already giving high friction, 1.4%/m is near the limit; go to 1.2–1.3%/m and watch friction force.
Rank 3 - change the flux to lower basicity (~1.0–1.1) and lower crystallisation temperature. A thinner liquid film and less crystal layer raise and homogenise the heat flux, giving a thicker shell at the meniscus and shallower oscillation marks. Side effects: higher heat flux and lower viscosity increase the risk of slag entrapment and of stickers if flux consumption falls; the flux must be re-tuned for the casting speed and the oscillation parameters, so it needs a trial rather than a same-shift swap.
3. Which action would you implement first and what would you monitor to distinguish success from a new problem?
Read the worked answer
First action: drop superheat to 20–25 K in the next tundish, in parallel with raising taper to 1.2–1.3%/m at the next mould change. Superheat is the free move - no lubrication risk, no trial needed - and 38 K is clearly too high for a 230 mm peritectic section. Keep the flux as-is for now so that two variables are not changed at once and the thermocouple map stays interpretable.
Markers of success (within 3–5 casts):
- Colder, more uniform mould thermocouple readings: the spread between the top-row and mid-row thermocouples, and between corner and mid-face, should fall noticeably; off-corner temperatures should come down toward mid-face values.
- Oscillation-mark depth decreasing from 0.9 mm toward 0.5–0.6 mm on a sulphur print of a sample slab.
- Slab inspection after scarfing shows a falling longitudinal-crack rate off-corner.
- Sticking alarms stop without any rise in friction force.
Markers of a new problem:
- Rising mould friction (withdrawal force or hydraulic pressure) or new sticking alarms: taper or flux is gripping, back taper off before breakout risk grows.
- Transverse or diagonal cracks, or a cold-shut: superheat is too low or casting speed is fluctuating.
- SEN blockage or stopper skulls: tundish temperature is being lost downstream of the superheat cut, so raise it 5 K and recheck.
- A change in the thermocouple pattern to a persistent single hot spot is a breakout precursor - stop, do not finish the sequence.