What 1.38 million sand tests actually show.
We went back through every reading from connected V-CAT II units over the five years to early 2025: about 5.6 million readings across 1.38 million test cycles. Each property was taken on the same sample in the same cycle, so these are honest pairs, not readings matched up afterwards.
The numbers.
| Property | Mean | Standard deviation | Valid readings |
|---|---|---|---|
| Compactability | 40.3 % | 3.3 | About 1.34 million |
| Moisture | 3.96 % | 0.72 | About 1.14 million |
| Green compression strength | 2.19 kg/cm² | 0.75 | About 1.11 million |
| Permeability | 136 AFS | 51 | About 146,000 |
There's no industry constant in this data.
Pool every plant together and the firmest relationship left standing is green strength against permeability, r = -0.27 on about 136,000 paired tests. Everything else averages out to nearly nothing.
A rule of thumb that holds across the industry doesn't exist in 1.38 million real tests, which is why generic plant software so often disappoints.
Now split the same test by molding line.
Four units carry enough paired readings to stand on their own. Moisture against compactability, on the same instrument model, running the same test on the same kind of green sand, gives a different answer on every one.
| Molding line | Correlation | Bar from minus one to plus one | Paired tests | What it means |
|---|---|---|---|---|
| Line A | +0.60 | Tens of thousands | Water drives compaction hard | |
| Line B | +0.28 | Hundreds of thousands | Present but muted | |
| Line C | +0.07 | Hundreds of thousands | Almost no relationship | |
| Line D | -0.42 | Tens of thousands | It runs the other way |
If your control rule came from a handbook, a seminar or another plant, this table is why it may not be working for you. The physics doesn't change between foundries. The muller, the return loop, the cooler, the ambient conditions and the way water is dosed all do, enough to reverse the direction.
Add the rejection record and it points at your defects.
Everything above links sand properties to each other. Add what was rejected, and the same arithmetic ranks sand behavior against your actual defects: which reading moved before the blowholes, and which one before the misruns.
Sand figures are measured, from connected V-CAT II units over the five years to early 2025. Readings outside physically valid ranges were excluded. The line comparison uses the four units carrying more than ten thousand paired readings; smaller samples are left out. Figures are rounded, foundries and lines are anonymized, and no customer, site or serial is identified.