Inoculation is one of the shortest-lived operations in a foundry. The effect starts fading within minutes of the addition, which means the timing matters as much as the product. Get it right and you control graphite structure and chill; get it wrong and you find out at the machining stage.
What inoculation actually does
Molten cast iron will solidify in one of two ways. It can follow the stable system and precipitate carbon as graphite, or it can follow the metastable system and form iron carbide — cementite — which is what foundries call chill. Chill is hard, brittle and effectively unmachinable.
Which route the iron takes depends on how many nucleation sites are available for graphite to form on. Clean, well-superheated iron has very few, so it undercools substantially before graphite can nucleate — and once undercooling passes a certain point, carbide forms instead.
An inoculant supplies those nucleation sites. It is not really an alloying addition: at 0.1–0.3% of the metal weight it changes the chemistry very little. What it does is seed the melt with fine particles that graphite can grow on, so solidification follows the graphite route at much lower undercooling.
The practical results are fewer chilled edges and thin sections, more and finer graphite, more uniform structure across varying section thicknesses, and — in ductile iron — a higher nodule count.
What is in an inoculant
Most commercial inoculants are ferro silicon based, typically around 65–75% silicon. The silicon is largely a carrier. The active work is done by small additions of other elements.
| Type | Active element | Best suited to |
|---|---|---|
| Calcium bearing | Ca, typically 0.5–2% | General purpose grey iron; the common baseline |
| Barium bearing | Ba, typically 1–3% | Where fade resistance matters — longer holding, longer pouring runs |
| Strontium bearing | Sr, typically 0.6–1.2% | Chill reduction in thin sections with minimal increase in section sensitivity |
| Zirconium bearing | Zr, often with Mn | Melts with nitrogen pinhole tendency; Zr helps fix nitrogen |
| Rare earth bearing | Ce, La | Potent nucleation; also helps neutralise trace subversive elements |
Trace aluminium and calcium are important in most of these: they form the oxide and sulphide particles that act as the actual nucleation substrate. An inoculant made from ferro silicon with essentially no aluminium or calcium is a poor inoculant, which is why "it's just ferro silicon" is a costly assumption.
Fade, and why timing dominates
Inoculation fades. The nuclei dissolve back into the melt or float out, and the effect decays — noticeably within 5 minutes, substantially by 10 to 15 minutes, and largely gone by 20.
This single fact drives most inoculation practice. The later you inoculate relative to pouring, the less inoculant you need and the more consistent the result.
Ladle inoculation
The alloy is added to the ladle during or just after filling, typically at 0.1–0.3%. It is simple, needs no special equipment, and is the most common method. It is also the most exposed to fade, so it suits short holding times and quick pouring.
Practical points: add into the metal stream rather than dumping onto the surface, size it appropriately for the ladle (roughly 0.2–6 mm depending on ladle size — the bigger the ladle, the coarser the material), and never add to a ladle that will then sit.
In-stream inoculation
A fine-grained inoculant, typically 0.2–0.7 mm, is metered into the pouring stream as the mould fills. Because the addition happens seconds before solidification, fade is almost eliminated and the addition rate can drop to 0.05–0.15%.
It gives excellent consistency and materially lower alloy consumption, at the cost of a dosing unit that has to be maintained and calibrated. For repetitive production work it usually pays for itself.
In-mould inoculation
A cast insert is placed in the running system and dissolves as metal flows past. Fade is effectively zero and addition rates are the lowest of the three, typically 0.05–0.1%.
The constraints are real: the running system has to be designed around it, the insert dissolution rate has to match the pouring rate, and any undissolved remnant is an inclusion. Best suited to high-volume, stable, repeated parts.
Grey iron and ductile iron are different problems
Grey iron
The aim is a uniform Type A flake distribution and no chill in thin sections. Watch for section sensitivity: over-inoculating heavy sections can coarsen the graphite and lose strength, so the addition should be tuned to the section thickness range in the casting, not just to the thinnest part.
Ductile iron
Inoculation follows magnesium treatment, and the order matters. Magnesium treatment strips sulphur and oxygen from the melt — the very elements whose compounds provide nucleation sites. Treated iron is therefore left badly nucleated and highly prone to chill, which is exactly why post-treatment inoculation is not optional in ductile iron practice.
Here the target is nodule count. More nodules means less segregation, better mechanical properties and a lower chill tendency. Late inoculation is especially valuable in ductile iron because the fade is compounded by the magnesium fade already underway.
Diagnosing problems
| Symptom | Likely causes |
|---|---|
| Chill in thin sections | Insufficient addition; fade from delayed pouring; excessive superheat; carbide-stabilising trace elements |
| Results vary ladle to ladle | Inconsistent holding time — the classic fade signature; inconsistent addition point |
| Low nodule count in ductile iron | Inoculation too early relative to pouring; magnesium fade; insufficient late addition |
| Slag or dross defects | Over-addition; oversized material not dissolving; poor slag removal after treatment |
| Pinholes | Nitrogen pickup; consider a zirconium-bearing grade; check for moisture |
Choosing, in practice
- Start with your method. Ladle, in-stream or in-mould determines the size fraction before anything else. The right chemistry in the wrong size will not work.
- Then your holding time. If metal sits before pouring, a barium-bearing grade will hold up better than a plain calcium one.
- Then your specific problem. Thin-section chill points to strontium. Pinholes point to zirconium. General grey iron work is usually fine on a calcium-bearing grade.
- Then trial it properly. Use chill wedge tests and metallography on your own iron. Inoculant performance depends on base iron chemistry, superheat and trace elements, so published comparisons only get you to a shortlist.
In short
Inoculation is a timing operation with a chemical component, not the other way round. Adding the right product too early will underperform a mediocre product added late. Before changing supplier, look at when the addition happens relative to pouring, how long the metal holds, and whether the size fraction matches the method — that is where most inconsistency lives.
We supply ferro silicon based foundry inoculants sized for ladle, in-stream and in-mould addition. Tell us your method and we will recommend a grade and arrange a trial quantity.
Frequently asked questions
What is fade in inoculation?
The gradual loss of inoculation effect over time as the nuclei that promote graphite formation dissolve or float out. Most of the effect is gone within 10 to 15 minutes, which is why late addition works better than early addition.
How much inoculant should I add?
Typically 0.1% to 0.3% of the metal weight for ladle inoculation, and considerably less for in-stream or in-mould methods because the addition happens immediately before solidification with almost no fade.