Casting defects · analysis

Heat-wise rejection analysis

Castings poured from one heat share the chemistry, the temperature and the treatment. Grouping defects by heat is how a foundry tells a melt problem from a mould problem, which is the difference between two entirely separate corrective actions.

By Versatile Equipments Pvt Ltd · equipment manufacturer & calibration provider · updated 2026-09-07

What the heat holds constant

Foundry analysis is mostly an argument about which subsystem is at fault: the melt, the sand, the tooling or the method. The useful cuts are the ones that hold most things constant so that one thing can vary.

Cavity-wise analysis holds the metal constant and varies the impression, which is why it points at tooling. Heat-wise analysis does the reverse. Every casting poured from one heat shares the same chemistry, the same treatment and roughly the same temperature, while passing through many different moulds. So a defect that clusters by heat is pointing away from the sand and the pattern, and toward the melt.

Which defects follow the heat

Some defect families are sensitive to what happened in the furnace and the ladle. These are the ones worth grouping by heat:

• Pinholes and gas porosity. Dissolved gas from a damp or rusty charge, a cold or wet ladle, or turbulence during handling. A single bad ladle can mark every casting it poured.
• Slag and dross inclusions. Carry-over from poor skimming, a dirty ladle or late treatment, so they travel with the metal rather than the mould.
• Misruns and cold shuts. Low tapping or pouring temperature, or metal held too long before pouring. Note these can equally come from gating and thin sections, so the heat pattern is what separates the two explanations.
• Chill and hard spots. Chemistry and inoculation. Where inoculation has faded, castings from the later part of a heat can machine harder than the earlier ones.
• Degenerate or low nodularity in SG iron. Magnesium treatment and its fade with holding time, so it is bound to the treated ladle and when the metal was poured from it.
• Shrinkage. Partly a feeding and method question, but chemistry and pouring temperature move it too, which is why shrinkage sometimes clusters by heat and sometimes by pattern.

By contrast, sand inclusion, scab, penetration, mismatch, flash and core shift generally spread across heats, because their causes sit in the sand system or the tooling. If one of those appears to cluster by heat, the more likely explanation is that something else changed at the same time, such as a sand batch or a shift changeover.

The multi-pattern test

An outlier heat on its own is suggestive rather than conclusive. The check that makes it convincing is simple:

• Did the defect affect several different patterns poured from that heat? If castings of three different parts, made on different plates and in different boxes, all show pinholes from the same heat, the shared element is the metal.
• Or was it confined to one pattern? Then the heat is probably a coincidence of timing, and the tooling, the method or that particular mould is the better suspect.

This is the same reasoning used at cavity level, applied one step up. Whichever factor is shared by the affected castings, and not shared by the unaffected ones, is where to look.

Position within the heat

A heat is not uniform from first mould to last. Temperature falls as the ladle empties, inoculation fades, magnesium fades in treated iron, and slag builds toward the end. Two useful patterns follow:

• Defects concentrated at the end of a heat suggest fade or falling temperature: chill and hard spots, degenerate nodularity, misruns and cold shuts in the last moulds poured.
• Defects concentrated at the start more often suggest ladle condition, such as a cold or damp ladle giving gas defects in the first castings before it warms through.

Seeing this requires the pouring order, or at least the time, to be recorded with the defect. Where only the heat number is kept, the effect still shows up as a bad heat, but the reason within it stays hidden.

Recording the heat so the analysis is possible

The heat number is usually the weakest link in a defect register, because it tends to be remembered at the end of the shift rather than read at the bench. A few practical points help:

• Use one unambiguous scheme. A date plus a sequence for that day is enough, and is hard to confuse across weeks. What matters is that the melt shop, the pouring log and the inspection bench all write it the same way.
• Capture it with the casting, not afterwards. A heat attributed hours later is the entry most likely to be wrong, and a wrong heat number quietly moves a defect onto an innocent melt.
• Treat it as the join key. The heat is what lets the defect register be read against the melt record and, through the time of pouring, against the sand lab record as well. Without it those records cannot be lined up at all.

From an outlier heat to an action

Once a heat stands out and the multi-pattern test supports it, the investigation moves to the melt record for that heat rather than to the moulding line:

• Chemistry as tapped and after treatment, against your own normal range for that grade.
• Tapping and pouring temperature, and how long the metal was held.
• Treatment and inoculation: quantity, method and the time between treatment and the last mould poured.
• Ladle condition: relined, dried, preheated, and how well it was skimmed.
• Charge for that heat: returns, rust, moisture or an unusual scrap batch.

Because these records already exist in most plants, the analysis is usually less about new measurement than about being able to point at the right heat with confidence. The value comes from the register carrying a heat number that can be trusted.

Heat-wise, cavity-wise and defect-wise cuts answer different questions and are strongest together, which is the subject of foundry rejection analysis. All three depend on the same foundation: a defect register where names, cavities and heats are recorded consistently at the bench.

Frequently asked questions