Ferrosilicon in Casting: Inoculation, Deoxidation and Cleanliness

Why Ferrosilicon Is Added to the Foundry Melt

In a foundry, ferrosilicon is not simply a source of silicon. It is a process additive whose value lies in three effects that occur at different moments in the melting and pouring cycle: deoxidation of the liquid metal, control of the solidification sequence, and nucleation of the graphite or carbide phases that decide the as-cast structure. Getting any one of those three wrong produces a casting that passes chemical analysis and fails inspection.

The distinction is important because ferrosilicon is often bought on silicon content alone. A lot of FeSi75 that is perfectly acceptable for steelmaking deoxidation can be a poor inoculant if its aluminium is high, if its sizing is wrong for the addition method, or if it has been stored damp. The specification that actually controls casting performance therefore includes the aluminium level, the level of any active element such as calcium, barium, strontium, zirconium or rare earths, the particle size distribution, and the freedom of the lot from moisture and slag contamination.

Deoxidation and Inclusion Removal in Iron and Steel Castings

Silicon dissolved in liquid iron or steel has a high affinity for oxygen. When ferrosilicon is added to the melt, silicon reduces dissolved oxygen and iron oxide according to reactions of the form:

Si + 2 O (dissolved) = SiO2

2 FeO + Si = 2 Fe + SiO2

The silica produced is less dense than the metal and, if the slag is fluid enough, it floats and joins the slag layer where it can be removed. In practice the reaction products must be liquid at the tapping and pouring temperature, otherwise solid silica particles remain suspended and can be trapped in the casting as non-metallic inclusions. This is why ferrosilicon is normally used in combination with other deoxidisers and with a slag system designed to absorb silica, and why the sequence of additions matters as much as the quantity.

The direct process benefit is a reduction in gas-related porosity. Oxygen dissolved in the melt participates in the carbon-oxygen reaction that produces carbon monoxide during solidification; lowering the oxygen activity before pouring removes the source of that reaction and reduces the pinhole and blowhole population in the casting surface and subsurface. Aluminium-bearing ferrosilicon adds a second, stronger deoxidation path through the aluminium, which is why a controlled aluminium level is sometimes deliberately specified in steel foundry practice, and why an uncontrolled aluminium level causes trouble in ductile iron.

Inoculation of Grey Iron: Graphite Morphology and Chill Control

The classic casting-performance role of ferrosilicon is inoculation of grey and ductile iron. Inoculation is a late addition of a silicon-bearing material to a melt that is already chemically correct, made immediately before or during pouring. Its purpose is not to change the analysis but to change the number and nature of the nuclei available for graphite growth.

Without effective inoculation, a grey iron melt of eutectic composition tends to solidify with undercooling. The consequences are a higher proportion of interdendritic or eutectic carbide, which appears in the microstructure as free cementite and in the casting as hard, unmachinable areas commonly described as chill or white iron in thin sections. This is a common failure mode in thin-walled castings, sharp corners and thin bosses, where the local cooling rate is high.

Inoculation with ferrosilicon raises the number of active nuclei, reduces the undercooling required for graphite nucleation, and promotes the growth of Type A flake graphite instead of undercooled Type D or Type E flake forms. The result is a more uniform structure, better machinability and more predictable mechanical properties. Graphite form in cast iron is evaluated under ISO 945-1 and ASTM A247, and Chinese foundries commonly apply GB/T 7216 for grey iron metallographic examination; the chill tendency itself is assessed by wedge chill testing under methods such as ASTM A367, in which the depth of white iron across a standard wedge is measured and used as a comparative index of inoculation effectiveness.

Ferrosilicon-based inoculant	Silicon base (%)	Active addition (typical, %)	Principal effect in the casting
Plain FeSi75	74.0-80.0	-	General nucleation and silicon addition; short-lived effect, sensitive to aluminium level
Calcium-bearing ferrosilicon	65-75	Ca about 0.5-2.0	Stronger nucleation, improved chill reduction in thin sections
Barium-bearing ferrosilicon	60-70	Ba about 2-6	Longer-lasting inoculation; suited to heavy sections and long holding times
Strontium-bearing ferrosilicon	65-75	Sr about 0.6-1.5	Reduced graphite degeneration and less risk of lumpy graphite formation
Rare-earth bearing ferrosilicon	45-75	RE about 0.5-2.5	Nucleation in melts with high residual trace elements, plus desulphurising assistance

Two practical rules follow from the chemistry. Strontium-bearing inoculants are preferred where graphite degeneration and lumpy graphite are the main risk, because strontium acts in a way that resists the deterioration of the spheroidal or flake form during long solidification times. Barium-bearing inoculants are preferred where the inoculation effect has to survive a long interval between treatment and solidification, because barium releases the nucleating effect more slowly than calcium. Neither is a substitute for correct base iron chemistry or for adequate superheat control.

Nodularisation and Post-Inoculation in Ductile Iron

In ductile iron the primary treatment is nodularisation with magnesium ferrosilicon, after which a separate post-inoculant is usually added to prevent the nodularity from decaying before the metal solidifies. Magnesium ferrosilicon is typically supplied with magnesium in the range of about 5-10% and rare earths at roughly 0.5-2%, on a silicon base near 44-48%, but the exact combination depends on the sulphur level of the base iron, on the treatment method and on the ductility requirements of the casting.

The ferrosilicon side of that treatment carries two specific risks. First, high-aluminium ferrosilicon in a ductile iron treatment increases the tendency to pinhole porosity and promotes dross and slag build-up, so low-aluminium material is specified for magnesium ferrosilicon and for the post-inoculant. Second, magnesium treatment generates a substantial amount of reaction product in the form of magnesium silicate and magnesium sulphide particles; if these are not removed through a clean slag, they appear as inclusions in the finished casting. Post-inoculation with ferrosilicon therefore does two things at once: it restores the nucleus population that the magnesium treatment destroyed, and it supports the flotation of the reaction products.

Nodularity itself is measured on polished sections under ISO 945-1 or ASTM A247, and the acceptance level is normally written into the casting specification; a minimum of 80-85% nodularity is a common contractual requirement for spheroidal graphite iron castings made to ISO 1083 or ASTM A536. The nodularity falls with every minute that passes between treatment and solidification, which is the direct reason why inoculant choice, addition method and pouring time are treated as a single process decision rather than three separate ones.

Addition Methods, Recovery and Fading

The same inoculant behaves differently depending on how it is introduced into the metal. The dominant variables are the residence time in the liquid, the degree of contact with the melt, and the amount of the addition that dissolves before the melt reaches the mould cavity.

Addition method	Typical addition rate	Where it is used	Main consideration
Ladle inoculation	About 0.2-0.6% of metal weight	General grey and ductile iron production	Strongest effect but subject to fading; addition rate varies with base sulphur and silicon
Stream inoculation	About 0.1-0.3%	Continuous pour and larger moulds	Bypasses the ladle, so less effect is lost before solidification
In-mould inoculation	About 0.1-0.3%	Thin-section and high-integrity castings	Effect delivered inside the mould; requires consistent granule sizing and mould design
Cored-wire injection	Set by wire design	Treatment of larger melts with tight control	Reproducible addition; wire fill weight and feeding rate must be verified per lot
Mould-height or mould-jacket addition	Grade dependent	Heavy sections where long solidification times demand persistent nucleation	Requires a slow-releasing inoculant grade, typically barium-bearing

Silicon recovery from a ferrosilicon addition in an induction furnace melt is generally high, on the order of 85-95%, because the addition is made into a well-mixed melt at a controlled temperature. The variable is not so much the recovery of silicon as it is the decay of the metallurgical effect. This is what the industry calls inoculation fade: the number of active nuclei falls with time, and the rate of decay depends on the inoculant's active element, the holding temperature and the time between addition and solidification. Ladle inoculation typically gives an effective window of roughly 5-15 minutes in normal foundry practice, which is why the pour should follow the addition without delay.

Granule size has to match the method. Powder and fine granule dissolve rapidly and are appropriate for stream and in-mould inoculation, where there is little time for dissolution; coarser granule or lump is appropriate for ladle additions where there is a short but real mixing period. Oversized material in an in-mould application can remain undissolved and produce a localised defect, while undersized material added to a ladle can be carried into the slag before it has any effect. Where cored wire is used, the fill weight and the wire composition must be within specification, because an under-filled wire silently changes the addition rate.

Defects That Ferrosilicon Practice Controls

Ferrosilicon influences a specific and recognisable set of casting defects. Mapping the defect back to the ferrosilicon decision is the fastest route to a fix.

Defect	Mechanism	Corrective action in ferrosilicon practice
Hard spots and chill in thin sections	Insufficient nuclei, undercooling, carbide stabilisation	Increase inoculation, use calcium or barium bearing grade, reduce time between addition and pour
Type D or Type E undercooled graphite	Late or weak inoculation in thin or rapidly cooled sections	Move to stream or in-mould inoculation; verify addition rate and granule size
Pinhole porosity in ductile iron	High aluminium in the inoculant or nodulariser base, combined with moisture	Specify low-aluminium ferrosilicon; store dry and avoid damp sand and wet charge materials
Slag inclusions and dross	Silica-rich reaction products not removed from the melt	Adjust deoxidation sequence and slag design; allow flotation time before pouring
Lumpy or exploded graphite	Excess rare earths or interference from trace elements	Reduce rare-earth addition; select a strontium bearing inoculant
Fading with long holding times	Loss of nucleation sites during holding and transport	Switch to a slow-release barium-bearing inoculant; shorten the holding interval

It is worth noting what ferrosilicon cannot do. It will not correct a base iron that has the wrong carbon equivalent, and it will not compensate for excessive superheat, a badly designed gating system or contaminated scrap. Inoculation works only when the base melt is already capable of solidifying with a sound structure; it changes the nucleation kinetics, not the thermodynamics of the alloy.

Specification, Verification and Handling of Foundry-Grade Ferrosilicon

For casting applications the purchase specification should be written around the effects being bought. A practical document states silicon as a range, caps aluminium to the level required by the process, sets the active element range for calcium, barium, strontium, zirconium, bismuth or rare earth bearing grades, names the particle size range together with permitted oversize and undersize, and requires a certificate of analysis per lot. Relevant standards are GB/T 2272 for ferrosilicon grades, GB/T 4333 series for chemical analysis, GB/T 13247 for particle size grading of ferroalloys, GB/T 3650 for packing and marking, GB/T 4010 or ISO 3713 for sampling, and GB/T 9441 for nodularity verification of the cast product.

Verification should include an incoming visual check for moisture, fines and slag, a size check by screening a representative sample, and periodic confirmation of the active element content. Where an inoculant is critical to a casting family, the most reliable control is a foundry-side trial: a wedge chill test or a stepped-section casting poured with the same treatment routine, sectioned and assessed against the required graphite form. That converts an abstract chemistry certificate into a measurable statement about how the material behaves in the actual mould.

Storage and handling deserve the same attention as in steelmaking supply. Ferrosilicon with roughly 30-90% silicon reacts with water and moisture to release hydrogen, and where phosphorus is present phosphine can also form, so material must be kept dry under cover on drained ground and enclosed spaces must be ventilated and gas-checked before entry. Moisture also causes the practical foundry problem of unreliable dissolution and an unpredictable inoculation response, which is a good reason to reject a wet lot on the basis of handling safety alone, before any metallurgical trial is run.

FAQ

Q: How does ferrosilicon improve casting performance?
It supplies nuclei that reduce undercooling during solidification, so grey iron grows a more uniform flake graphite structure instead of carbides, and the casting develops better strength, hardness and wear resistance with less chill in thin sections. It also lowers the oxygen activity of the melt, which reduces gas porosity.

Q: How does ferrosilicon reduce impurities in the melt?
Silicon combines with dissolved oxygen and iron oxide to form silica, which floats into the slag and can be removed before pouring. The reduction in dissolved oxygen also suppresses the carbon-oxygen reaction that generates blowholes, leaving a cleaner and more uniform casting with more consistent mechanical properties.

Q: What addition rate of ferrosilicon inoculant is normal?
Ladle inoculation is typically about 0.2-0.6% of the metal weight, while stream and in-mould inoculation are typically about 0.1-0.3%. The correct rate depends on the base iron sulphur and silicon levels, the section thickness and the pouring time, and it should be set by sectioning trial castings rather than by a single rule.

Q: What is inoculation fade and how is it controlled?
Fade is the loss of active nucleation sites during holding before solidification, and it reduces the graphite control that the inoculant was added to produce. Ladle inoculation typically remains effective for about 5-15 minutes. Barium-bearing inoculants fade more slowly and are used for heavy sections, long holding times and in-mould applications.

Q: Why does aluminium content in ferrosilicon matter in ductile iron?
Aluminium is a strong oxide former and, at the levels found in high-aluminium ferrosilicon grades, increases pinhole porosity, dross and slag build-up in magnesium-treated iron. Ductile iron producers therefore specify low-aluminium ferrosilicon for both the nodulariser base and the post-inoculant.

Q: How is the effect of an inoculant verified in production?
Chemical verification uses the GB/T 4333 series for analysis with sampling under GB/T 4010 or ISO 3713, and particle size is checked under GB/T 13247. Metallurgical verification uses graphite form assessment under ISO 945-1 or ASTM A247, nodularity checking under GB/T 9441 for ductile iron, and wedge chill testing under methods such as ASTM A367 for chill tendency.