The pour that should have been stopped: wet floor, wet ladle, dead extractor
In a foundry pouring grey iron from a 500 kg induction furnace, a 90 kg mould was closed at 09:00 and is due to be poured at 15:00. The mould is 400 × 400 × 250 mm in green sand at 6.5% moisture, with a linseed-oil-bonded core that has been stored uncovered. The floor beside the mould is still wet from the previous shift's washing, and a small spill on the previous pour 'spat' metal. The ladle lining was patched with wet refractory an hour ago and has not been dried. Two workers wheel the ladle by hand; one wears leather shoes without spats, the other has no face shield, only sunglasses. The pouring bay extractor has been out of order for a week and the area smells strongly of burnt core binder. Pouring temperature is 1420 °C. The supervisor wants to pour today to meet the despatch date.
1. List the hazards in this situation and rank them by the severity of the likely consequence, not by how obvious they are.
Read the worked answer
Ranked by severity:
- Water in contact with molten metal - steam explosion. Wet floor, an undried patched ladle lining and a green sand mould at 6.5% moisture closed for six hours are three independent water sources. A litre of water flashes to roughly 1 700 litres of steam; the expansion throws metal 5–10 m. The earlier 'spitting' was a small warning of exactly this mechanism. Potential consequence: multiple fatalities.
- Confined gas from the core plus a dead extractor. The oil-bonded core and the mould continue to evolve hydrocarbons and CO after pouring, and a closed mould with unvented core gas can blow metal out of the sprue or the parting line. With no ventilation, the pouring crew is also breathing CO and binder decomposition products in an enclosed bay.
- Handling a 90 kg hand-wheeled ladle. Two people pushing a ladle of 1420 °C metal have no control if a wheel catches or the metal surges; spillage onto legs and feet is the classic foundry burn.
- PPE failure. No face shield, no spats, open leather shoes - the exact areas where spills land. Sunglasses give no impact or splash protection.
Items 1 and 2 are process-level hazards that no amount of care by the pourer can control; items 3 and 4 are behaviour and equipment issues that add to the consequence.
2. State whether you would allow the pour, and set out the controls in the order of the hierarchy of control.
Read the worked answer
Do not pour. The water-related hazards cannot be removed by care at the ladle; the risk cannot be brought to an acceptable level within the shift, so the production pressure must not decide it.
Hierarchy of control:
- Elimination / postponement: stop the pour, reschedule. A 90 kg mould can wait; it can also be stripped and re-made if the sand is too wet.
- Engineering controls: dry the floor with dry sand or absorbent and keep the pouring bay dry; dry the patched ladle with a controlled burner and never fill a wet or cold ladle; connect core vents and mould vents and vent the cope; repair or replace the extractor, or provide forced local ventilation before any pour; provide a mechanical ladle hoist or bail so the ladle is not hand-wheeled.
- Administrative controls: a written permit-to-pour that checks ladle dryness, floor condition, ventilation and PPE before each heat; a marked exclusion zone with only the pourer and one helper; training in the steam-explosion mechanism, including the fact that a wet patch under a mould is a fatal hazard, not a nuisance.
- PPE (last line): face shield plus foundry goggles, aluminised apron and leggings, spats, heat-resistant gloves, safety shoes.
The order matters: issuing a face shield and pouring anyway would satisfy none of the hierarchy and would leave the fatal hazard untouched.
3. The supervisor argues the core has been 'baked, so it is dry' and the mould has been closed for six hours, 'so it must be dry by now'. Answer that argument with the relevant physics.
Read the worked answer
Both statements are wrong, for the same reason: a closed mould does not dry, it traps water.
A green sand mould at 6.5% moisture contains roughly 6.5% of its mass as water. A mould of that size holds several kilograms of water, and once the box is closed the only escape is through the sand to the atmosphere, which is slow at ambient temperature and effectively stops under the cope. Six hours of standing does not remove the moisture from the mould wall next to the cavity; it may even redistribute it towards the cooler cope. When the metal arrives, that water flashes to steam and the expansion drives metal back out.
A baked oil-sand core is not dry either. It was baked at 200 °C, which drives off most solvent, but it has then been stored uncovered for 10 days in humid air and re-absorbed moisture. Its vents were not connected, so the steam and binder gases generated on contact with 1420 °C metal have nowhere to go except through the metal.
The correct test is not 'it looks dry' or 'it has stood for hours'; it is a measured moisture check on the sand and a check that the core is still warm and dry from the oven, plus proven vent paths. Without those, the pour is a gamble with a known fatal failure mode.