Ask three foundries what FeSiMg8RE3 means and you will usually get three partial answers: “8% magnesium”, “a grade”, “the one we have bought for years”.
Here is the answer in one line, and everything below is elaboration:
A nodularizer grade code is not a model number. It is a compressed sentence. Read it in three parts — FeSi is the carrier, Mg8 is the working element at roughly 8%, and RE3 is roughly 3% rare earth, which is an insurance element and has almost nothing to do with how strongly the alloy nodularizes.
That third part is the one most buyers skip, and it is where both the cost and the risk sit. It is also the part the grade code tells you the least about: whether those rare earths are cerium and lanthanum or yttrium-based decides whether the grade resists interference or resists fading, and those are two different jobs.
Seven questions follow, in the order they tend to arrive.
Start with the three-part split, then with what the code leaves out.
The carrier — FeSi. Magnesium boils at about 1091°C while nodularizing treatment runs at roughly 1380–1520°C. Pure magnesium dropped into iron at that temperature flashes off and the reaction is violent. Diluting it in ferrosilicon is what lets it release steadily instead. So the silicon in a nodularizer is primarily a carrier, not an active ingredient — the same point made in the inoculant comparison, and worth repeating because it is the most common misreading.
The working element — Mg8. Roughly 8% magnesium. This is the number that sets your addition rate, and it is a nominal figure.
The insurance — RE3. Roughly 3% rare earth. Section 2 covers what it does; the short version is that it protects the magnesium rather than adding to it.
Whatever comes after. Some producers append symbols for additional elements (Ca, Ba among them). What a given symbol means in a given producer’s recipe is not universal — ask rather than assume.
Now the more useful table. Most buying problems come from the second column:
|
|
What the grade code gives you |
What the code does not give you |
|
Carrier |
Ferrosilicon base |
Silicon range, and whether your melt has room for the silicon it carries |
|
Working element |
Mg, as a nominal percentage |
Whether that is total Mg or effective Mg; the tolerance; the test method |
|
Oxidised Mg |
— |
The MgO ceiling — magnesium already locked in oxide does no nodularizing work |
|
Rare earth |
RE, as a nominal percentage |
The split between light (Ce / La) and heavy, yttrium-based RE |
|
Fading resistance |
— |
Whether any yttrium-based heavy RE is present |
|
Additional elements |
Sometimes a symbol |
What that element is there to do, and whether you need it |
|
Particle size |
Never |
The size band, plus oversize and fines limits |
|
Addition rate |
Never |
Derived from your recovery — not a property of the grade |
|
Batch traceability |
— |
Certificate of analysis per batch, traceable by heat number |
Two conclusions follow from that second column, and they are the practical value of this section:
One: the same code from two suppliers is not the same product. Nothing in “FeSiMg8RE3” fixes the MgO ceiling, the rare earth split, the particle size band or the tolerance. Two materials can both honestly carry that code and behave differently in your ladle.
Two: therefore the grade code is a starting point for the conversation, not the specification. Everything that actually governs behaviour has to be agreed separately and in writing. Section 8 sets out that list.
They neutralise interference. That is the job, and it is not the same job as nodularizing.
Titanium, lead, bismuth, antimony and arsenic in the base iron degrade graphite nodule roundness. Rare earths bind with those elements preferentially, which leaves the magnesium free to get on with spheroidising. Without that protection, part of your magnesium is spent fighting interference instead of doing the work you paid it for.
So the useful mental model is:
Rare earth is insurance, not nutrition. It does not make the nodularizer stronger. It makes the nodularizer’s magnesium go further on dirty iron.
Three consequences that follow directly.
Insurance has a premium. Rare earths are not free, and they are not inert: excess rare earth raises the tendency to chill and cementite, particularly in thin sections. If your charge is clean and your recovery is good, a high-RE grade is paying for cover you do not need and picking up a side effect you did not want.
There are two kinds of cover, and the code does not say which you are buying.
|
|
Light rare earth (Ce / La based) |
Heavy rare earth (yttrium based) |
|
Primary job |
Neutralising interfering elements |
Resisting fading over long solidification times |
|
Where it earns its keep |
Variable scrap, cupola iron, dirty charge |
Heavy sections, slow pours, long holding |
|
What it does not solve |
Fading |
Interference |
|
Typical RE band direction |
RE 1%–5% depending on charge cleanliness |
Specified for the casting, not chosen by habit |
The distinction matters because the two get confused constantly. A foundry with fading problems on thick castings raises RE, sees no improvement, and concludes rare earths do not work — when what it needed was yttrium-based heavy RE, or a change to inoculation timing. Conversely, a foundry with variable scrap buys a yttrium grade and pays for anti-fading capacity it does not draw on.
The band follows your charge, not your habit. Clean electric-furnace charge with stable, low sulphur wants the low end. Variable scrap sources want more. Cupola iron, higher in both sulphur and interference, wants more again. If raising RE does not improve nodularity but does raise chill, the constraint is somewhere else — go back to sulphur and temperature before touching the grade again.
Six items, in the order they tend to cost money.
|
Item |
Why it is on the sheet |
What to agree in writing |
|
Mg — total vs effective |
Only part of the magnesium is available to nodularize |
Which basis the quoted Mg uses, and the tolerance |
|
MgO |
Magnesium already present as oxide does no nodularizing work |
A ceiling. This is the usual place a cheap grade hides its discount |
|
RE split |
Light and heavy rare earth do different jobs (section 2) |
Total RE, plus the Ce / La / Y split where it matters |
|
Si |
The carrier brings silicon into the melt |
Range, and how it fits your final silicon budget |
|
Al |
Carried in with the ferrosilicon base; normally held low |
A ceiling if your castings are sensitive to it |
|
Particle size and distribution |
Governs dissolution speed, and therefore recovery |
The band, plus oversize and fines limits |
MgO deserves a note of its own, because it is the single most common source of a “cheaper” quotation that is not cheaper. Two grades can both be quoted at 8% Mg and behave differently, because total magnesium and available magnesium are different numbers. Magnesium tied up as MgO is billed at magnesium prices and does nothing. Comparing grades on Mg content alone, without an MgO ceiling, is comparing the wrong number.
Particle size is the second one that gets treated as a packing detail rather than a process variable. Size governs how fast the alloy dissolves, which governs how much magnesium flashes off before it can do anything — which is why the same grade at two size bands gives two different recoveries. Match the band to the addition method rather than accepting whatever ships.
Resist the temptation to look for a casting-to-grade lookup table. There is no honest one, because the grade is not determined by what you are casting — it is determined by three conditions.
Condition one: your treatment process, because it sets recovery. Recovery is the divisor in every addition calculation, so it moves the required Mg band more than anything else. Broadly, the more controlled the process, the lower the Mg band it suits — because the point of a controlled process is to waste less magnesium, and you then do not need a high-Mg grade to land the same residual.
Condition two: base iron sulphur and interference. Sulphur consumes magnesium in a fixed, calculable way; interference consumes it indirectly, through rare earth demand. Cupola iron and variable scrap push both up.
Condition three: how long the magnesium has to last. This is the axis that maps most closely onto casting type, and it is the one worth thinking in. What distinguishes a thin-wall casting from a heavy section is not the alloy, it is how much time passes between treatment and the last of the iron solidifying. Magnesium fades across that interval. Short interval, little fading, modest requirement. Long interval on a heavy section, and the question is not how much magnesium to charge but whether any is still there at the end.
Read against those three, the directions look like this:
|
Situation |
Grade direction |
Reason |
|
Open ladle / pour-over, cupola iron |
Higher Mg band, mid-to-high RE |
Low recovery and high sulphur both push consumption up |
|
Covered ladle or sandwich, electric furnace, stable low sulphur |
Mid Mg band, low RE |
Recovery is decent and the charge is clean |
|
Transfer ladle (GF / rocking), electric furnace |
Lower Mg band, low RE |
High recovery means less Mg is needed to land the target |
|
Cored-wire injection |
Lower to mid Mg band |
Recovery depends on injection depth and speed; size requirements are tighter |
|
In-mould nodularization |
Lower Mg band, tightest size control |
The reaction chamber design governs everything else |
|
Thin sections, fast pour |
Lower Mg band, RE at the low end |
Guard against chill and cementite |
|
Medium sections (roughly 15–50 mm) |
Mid band |
The normal selection window |
|
Heavy sections (>80 mm), slow solidification |
Yttrium-based heavy RE |
The requirement is fading resistance, not more magnesium |
|
Variable scrap, unknown interference |
Mid-to-high RE |
Budget the insurance |
|
Long holding or long pouring rhythm |
Heavy RE plus late inoculation |
Compress the interval, or pay for fade resistance |
The question nobody asks, and the one that separates a specification from a habit. Four situations where the answer is at least “not in this form”.
One: your final silicon has no headroom. The carrier is ferrosilicon, so every kilogram of nodularizer carries silicon into the iron. Where the silicon budget is genuinely tight, a low-silicon or silicon-free magnesium source deserves a conversation with your supplier — the alternatives exist (nickel-magnesium, pure-magnesium cored wire among them) and they are priced accordingly.
Two: your charge is genuinely clean and your recovery is genuinely high. Low stable sulphur, few interfering elements, a controlled treatment process. Under those conditions a high-RE grade is insurance you are unlikely to claim on, and it brings a chill tendency you did not need. The correct move is to move down the RE band and spend the difference on recovery instead.
Three: the real problem is fading, and you are buying more RE to fix it. This is the error that burns the most trial time. Fading is about time and solidification, not about interference. More cerium will not fix it; yttrium-based heavy RE, late inoculation, or compressing the interval between treatment and pouring will. Before ordering a different grade, check which of those three you actually have.
Four: your returns make up a large share of the charge. One factor that is easy to overlook: rare earths are not lost from the melt as readily as magnesium is. Where a high-RE grade is used continuously and returns are recycled at a high ratio, it is worth tracking whether RE is accumulating in the circulating material, and setting the RE band with that in mind rather than by habit. Check it against your own returns ratio — the answer differs between foundries.
None of these mean “do not use FeSiMg”. They mean the grade direction has to be argued from your conditions, and that a supplier who only ever offers what you ordered last time is not arguing at all.
It is an output, not an input — and it is not a property of the grade.
Any addition rate quoted without your recovery, your base sulphur and your target residual magnesium attached is a number borrowed from somebody else’s process. This piece deliberately does not print one. The derivation is short and it is yours to run:
Mg consumed by desulphurization = (S before treatment − target S) × 0.76
Total Mg required = (target residual Mg + desulphurization Mg) ÷ Mg recovery
Nodularizer addition = total Mg required ÷ Mg content of the grade
The 0.76 is the ratio of atomic weights (Mg 24.3 ÷ S 32.06). Typical residual magnesium sits around 0.03%–0.05%: below that, nodularity suffers; much above 0.06%, chill and cementite risk enters.
Three things that follow and are worth holding onto:
The grade code covers perhaps a third of what actually matters. The rest has to be in the order.
|
Clause |
Why it is there |
|
Mg content, with the basis stated (total or effective) and a tolerance |
Without the basis, two suppliers quote the same number and mean different things |
|
MgO ceiling |
The place where “the same grade” stops being the same product |
|
Total RE, plus the Ce / La / Y split where it matters |
Light and heavy rare earth do different jobs |
|
Si range |
Final silicon budget |
|
Ceiling on the elements you care about (Al among them) |
Depends on your castings |
|
Size band, with oversize and fines limits |
Governs dissolution speed and therefore recovery |
|
Test method and the standard it refers to |
Decides whose number wins in a dispute |
|
Certificate of analysis per batch, traceable by heat number |
Turns a complaint into a checkable fact |
|
Packing and moisture protection |
Magnesium-bearing alloy degrades in damp storage |
One place to see what the grade code carries and what it cannot carry.
|
Parameter |
Does the grade code tell you? |
Must be agreed separately |
Covered in |
|
Carrier metal |
Yes — FeSi |
Si range, and your silicon headroom |
Q1, Q3 |
|
Working element |
Yes — nominal Mg |
Total vs effective Mg, tolerance, test method |
Q1, Q3 |
|
Oxidised magnesium |
No |
MgO ceiling |
Q3 |
|
Rare earth total |
Yes — nominal RE |
Tolerance |
Q2 |
|
Rare earth type |
No |
Light (Ce / La) vs heavy (Y) split |
Q2 |
|
Fading resistance |
No |
Yttrium-based heavy RE where needed |
Q2, Q4 |
|
Additional elements |
Sometimes a symbol |
What each is for, and its range |
Q1 |
|
Particle size |
No |
Band, plus oversize and fines limits |
Q3, Q8 |
|
Addition rate |
No |
Derived from recovery — not a grade property |
Q7 |
|
Grade direction |
No |
From process, charge and time-to-solidification |
Q4, Q6 |
|
Batch traceability |
No |
COA per batch, heat number |
Q8 |
Henan Longchuang Metallurgical Material Co., Ltd. supplies steel mills and foundries with RE-MgFeSi nodularizers, inoculants, alloy cored wire, ferrosilicon, manganese and chromium series alloys and some 20 further products, produced to specified composition windows and particle size requirements.
On grades specifically:
Tell us your section thickness, treatment process, base iron sulphur and how long the iron has to last before it solidifies — grades available, recommended direction and quotation back to you within 【12小时】.