Green sand system out of balance: diagnosis from the sand laboratory sheet
At a jobbing foundry casting 2–8 kg grey iron valve bodies in green sand, the sand plant mixes 1 000 kg batches of silica sand (AFS 55), bentonite, water and a little sea-coal. Over three weeks the rejection rate for blows and rough surface rose from 3% to 11%. The moulder complains that the sand balls in his hand and sticks to the rammer, the cope is soft, and patterns are hard to draw. The laboratory reports on a hopper sample: moisture 6.8%, clay by AFS wash test 8.9%, green compressive strength 0.55 kg/cm², permeability 62, shatter index 68, loss on ignition 3.2%. The specification for this work is moisture 4.5–5.5%, clay 11–13%, green strength 0.8–1.1 kg/cm², permeability 120–160, LOI 2.0–3.0%. There is no sand cooler, and mixed sand is used at 45 °C. Only bentonite and water are added at the muller, which runs a 90-second cycle.
1. Interpret the test sheet. Which properties are out of control, and what single physical condition explains the low strength together with the low permeability?
Read the worked answer
Compare each reading with specification:
| Property | Actual | Specified | Verdict |
|---|---|---|---|
| Moisture | 6.8% | 4.5–5.5% | high |
| Clay (AFS) | 8.9% | 11–13% | low |
| Green strength | 0.55 kg/cm² | 0.8–1.1 | low |
| Permeability | 62 | 120–160 | very low |
| LOI | 3.2% | 2.0–3.0% | high |
The controlling number is the water-to-clay ratio: 6.8 / 8.9 = 0.76, against a correct ratio of about 5.0 / 12 = 0.42. There is too little clay to hold too much water, so water is present as free water in the pores. Free water blocks the interstices, which is why permeability has collapsed to 62, and it dilutes the clay bridges, which is why green strength is only 0.55 kg/cm². Wet sand is also why the moulder's sand balls, sticks to the rammer and draws badly. The high LOI shows over-adding of sea-coal (plus dead clay and fines), which further closes the voids. This is not one bad property but one wrong balance, and adding water or sea-coal would make it worse.
2. The plant engineer proposes to correct the batch by adding water only, 'because the sand looks dry'. Give a corrected addition plan per 1 000 kg batch and state the mixing and temperature changes needed.
Read the worked answer
Adding water is the opposite of what is needed; the sand looks dry because it is hot and the free water evaporates at the surface.
Corrective additions per 1 000 kg batch:
- Bentonite: raise clay from 8.9% to 12% → 0.031 × 1 000 ≈ 31 kg, say 35 kg to cover handling loss.
- Water: bring moisture down towards 5.0%, i.e. remove about 18 kg of water per tonne. But the new clay absorbs roughly a third of its mass in water (≈11 kg), so cut water in stages and re-test rather than calculating once. Start by adding the clay and no water, mull, sample, and add small increments only if the moisture falls below 4.5%.
- New sand: add 5–10% fresh sand per batch to dilute accumulated fines and dead clay, since there is no way to remove them.
- Sea-coal: reduce the addition to bring LOI back to ≤3.0%.
Process changes: the 90-second muller cycle cannot wet and activate 31 kg of bentonite. Extend to 4–6 minutes, and mull after adding clay before final water adjustment. At 45 °C the bentonite hydrates poorly and the moisture reading drifts, so add a sand cooler or allow the heap to cool before mixing. Re-test moisture, green strength and permeability every shift until three consecutive samples are in specification.
3. Explain why the defects produced (blows, rough surface) follow from these properties, and why simply increasing ramming pressure would not cure them.
Read the worked answer
Blows and rough surface are gas defects, and the gas has to escape through the mould wall or not at all.
With permeability 62 instead of 120–160, gas generated at the metal front (steam from free water, volatiles from sea-coal) cannot pass through the sand fast enough. Pressure builds at the mould–metal interface; where the skin is still weak the gas pushes it back, producing blows just under the surface and a rough, scabby skin. The low green strength (0.55 kg/cm²) makes it easier for the gas to lift the sand, and the wet sand sticks to the pattern, tearing the surface and leaving a rough cavity.
Ramming harder is the wrong control because it reduces permeability further by closing the voids, and it increases the risk of spring-back and draw tears. Compaction cannot compensate for a sand with no void network. The correct sequence is to fix the sand first (clay up, free water down, longer mulling), then set ramming to give a uniformly hard mould at a sensible permeability. Venting the mould with a vent wire helps as a temporary measure, but the permanent cure is the water-to-clay balance.