Carbides in ductile iron are a cooling and chemistry problem that MgFeSi can either cause or cure. When magnesium or rare earth is at the top of the grade window, when silicon is low, or when inoculation is late or insufficient, thin sections chill white and machining becomes difficult. Thai automotive foundries that hold carbide free thin walls do three things together: they keep magnesium residual inside 0.035 to 0.055 percent, they hold carbon equivalent near 4.3 to 4.5 for thin sections, and they inoculate at the pouring stream or in the mould rather than only in the ladle.

Thin wall automotive parts fail first at the extremities, where cooling is fastest. A brake caliper bracket or a thin housing rib can show a chill depth of a millimetre under the skin while the section centre is fully nodular. That chill is not a moulding problem, and adjusting sand or pouring temperature alone will not remove it if the alloy and inoculation balance is wrong.

Magnesium and rare earth are carbide forming in the sense that they raise the tendency to chill, especially when the iron is short of silicon or nucleation.

The mechanism is not that magnesium is a carbide former in the alloying sense, but that a strong nodulizing addition reduces the graphite nuclei population unless inoculation restores it. The stronger the nodulizing effect, the more nucleation has to be supplied afterwards.

Grade selection decisions made for heavy sections become carbide risks in thin ones.

The cleanest rule for thin wall automotive work is to use the lowest grade that reliably covers the sulfur load, with rare earth at 0.5 to 1.5 percent, and to buy silicon at the upper half of the 38 to 44 percent range rather than the lower half.

Carbon and silicon set the graphitizing potential that MgFeSi works against.

Silicon deserves separate attention because MgFeSi is itself a silicon carrier. A grade with 38 percent silicon delivers less graphitizing silicon per tonne than one at 44 percent, so changing grades changes silicon balance as well as magnesium, and the charge calculation has to account for both.

Inoculation is the counterweight to the nodulizing addition, and its timing matters as much as its quantity.

1. Treat with MgFeSi in the covered ladle and let the boil finish fully

2. Add a ladle inoculant only as a base addition, not as the main inoculation

3. Add stream inoculation during pouring for thin wall moulds

4. Use in-mould inoculation where the thinnest sections sit far from the sprue

5. Verify chill depth on a wedge sample from the first mould of each pattern

6. Re-check after any change of MgFeSi grade, size band or supplier lot

A foundry that changes grade without re-checking the wedge is testing the change on production castings rather than on a sample.

Carbides and low nodularity look similar in a photograph and behave differently in a test.

When both signals appear, the correct order of correction is inoculation first, then alloy weight, then carbon equivalent, because inoculation is the cheapest and fastest change to test.

A sequence that resolves most carbide problems without disrupting production.

• Confirm final silicon and carbon equivalent on the last three heats rather than the nominal charge

• Review the MgFeSi grade against the section range, not against the previous pattern

• Add or increase stream inoculation on the thinnest patterns

• Reduce magnesium residual to the lower half of the window where sulfur allows

• Check wedge chill depth per pattern and record it with the heat number

• Revisit pouring temperature, since colder iron increases chill independently of chemistry

The last item is the one most often misread, because cold iron produces a carbide defect that chemistry cannot fix.

It can raise carbide tendency when magnesium or rare earth is at the upper end of the window and inoculation does not restore nucleation. The alloy is one contributor among several, not the only cause.

A residual of about 0.035 to 0.045 percent is commonly used for thin sections, while heavier sections can run higher. The safe value is the lowest figure that still covers sulfur and holds nodularity.

It generally increases carbide tendency in thin sections while reducing graphite degeneration in heavy ones. That is why rare earth should follow section thickness rather than be increased for safety.

Often it can, which is why inoculation is the first trial. If inoculation does not move the chill depth, the alloy weight or grade is the next suspect.

Because cooling rate is highest there. Chemistry sets the tendency, and the edge is where that tendency is expressed first.

It should be. Silicon from 38 to 44 percent in the alloy is a real input to final silicon, and changing grades without adjusting the charge shifts carbon equivalent.

A wedge or step bar poured from the same iron and broken after cooling shows the chilled zone. It is a shop floor test that costs one mould per pattern.

Increasing the MgFeSi weight. That raises magnesium, adds more chill tendency, and usually makes the defect worse while adding cost.

ZhenAn supplies FeSiMg in grades from FeSiMg6Re1 to FeSiMg10Re7 with silicon in the 38 to 44 percent range, in 10 to 50 mm and 50 to 100 mm lump plus granule form, ISO 9001:2015 certified, packed in 1 MT big bags at 20 to 25 MT per 20 foot container.

Share the section thickness, base iron sulfur and inoculation practice to have a grade recommended for carbide control.

Silicon and rare earth ranges relevant to carbide control are published on the 38 percent silicon FeSiMg grade table page and the FeSiMg6Re1 nodulizer page. For thin wall automotive work the finer band on the 3-6 mm FeSiMg granule page reacts quickly enough for in-mould use.