Magnetic Silicone Molding: Incorporating Ferrite Powders into RTV Bases

Magnetic silicone is polysiloxane loaded with ferrite until the cured part answers a magnet. Pull scales with how much ferrite you actually disperse. Mixability scales the other way.

RTV cups hit the wall first: dense powder, agglomerates, settling, and a paste that will not wet a master. Two-roll HCR mills take more ferrite because shear wets. LSR magnetic grades are a specialist fill, not a bucket experiment. Do not invent a loading. Write the function (hold this weight on this geometry, close this switch at this gap) and let the compounder pick a ferrite and a process that still mixes.

What “magnetic silicone” is (and is not)

Ferrites used in rubber are ceramic powders: magnetite (Fe3O4) and related iron oxides, hard ferrites (barium or strontium ferrite) for permanent-magnet compounds, and some soft MnZn / NiZn ferrites when the part is a flexible flux path rather than a refrigerator magnet. They are not interchangeable with carbon-black “conductive” grades, with carbonyl iron for EMI, or with a fridge-magnet sheet that is actually a filled thermoplastic.

Unfilled silicone is not magnetic. A pinch of ferrite that tints the rubber grey will not hold a tool to a press. Magnetic moment is a volume-fraction problem. Below a useful packing, you have a dirty insulator. Across it, you have a heavy, stiff, abrasive compound whose pull still depends on thickness, on whether the powder was aligned in a field during cure, and on the magnet you are pairing with.

Color goes dark. Density jumps. Elongation and tear drop because ceramic does not stretch. Compression set gets worse if you pack the gum to the edge of mixability. None of those are reasons to skip ferrite if the part must locate on a steel fixture. They are reasons not to use magnetic RTV as a general mold rubber.

This article is ferrite in the silicone, often to mold magnetic parts, gaskets with locating pull, toys, sensors, or fixtures. It is not about magnetizing a steel mold.

Ferrite loading versus mixability

The trade is blunt:

• More ferrite volume: more remanence or permeability (family-dependent), more weight, more viscosity, worse tear, more abrasion, more settling if the mix sits.
• Less ferrite: it mixes, it pours, it looks like silicone, and it barely holds a paperclip.

RTV (platinum addition-cure or tin condensation-cure) is the least forgiving carrier. Low shear in a cup does not wet a dry ferrite bed. You get dry pockets that are magnetic crumbs in a soft matrix, which is a weak magnet and a tear starter. High loading makes a non-pourable paste. That paste still settles: dense particles fall, polymer rises, and the first pour from the cup is ferrite-starved.

Practical mixing control without a fake percent:

• Dry-blend or masterbatch the ferrite into one component only if the supplier says that component can take it. Dumping powder into mixed A+B wastes pot life.
• Wet the powder. Unwetted ferrite is air. Vacuum after wetting, not instead of wetting. Vacuum on a dry powder dump expands air and can foam the cup.
• Use a thixotropic base (treated fumed silica already in the RTV) if you need the suspension to hold through gel. A water-thin platinum will stratify on a 30-minute pot life.
• Do not chase loading past the point you can scrape a homogeneous streak on a card. Streaks are loading gradients. They are also air.

HCR milling is how production flexible magnets get made. The mill applies shear the cup cannot. Addition-cure or peroxide HCR magnetic lines exist as compounded grades. That is the path when the RFQ needs repeatable pull on a gasket.

LSR: only if it is an injectable magnetic LSR from a compounder. You will need larger gates, abrasion-aware steel, and a fill pattern that does not powder-starve a weld line (a ferrite-poor knit is a dead magnetic zone).

Settling, orientation, and platinum cure

Ferrite density is several times the gum. In a still RTV cup the powder falls. In a tall mold the bottom becomes the magnet and the top becomes rubber. Remedies are process, not slogans: thixotrope, mix immediately before pour, shorter gel, sometimes a rotating fixture for a prototype, and inspection of density or of pull on coupons from top and bottom of the same shot. A single “it stuck to the fridge” demo on the cup leftover is not a part spec.

Some magnetic parts are anisotropic: a field during cure orients hard-ferrite platelets or domains so pull is stronger through the thickness. That is a production setup (coil, fixture, cure in field). A random RTV pour is isotropic-ish and weaker for the same loading. If the drawing shows a pole face, ask whether orientation is required. Shop RTV will not invent it.

Platinum RTV is picky. Some iron-oxide and ferrite packages, some coatings, and some dirty mill-scale powders interfere with addition cure (sticky faces, under-cure in the sludge layer). Magnetite is also a common black pigment in silicone, which does not make every ferrite “platinum safe.” Test the actual powder in the actual base on a plaque, including the heavily loaded bottom of a settled cup. Tin-cure is less catalyst-sensitive and still a poor library tool (it keeps shrinking). Do not pick tin only to dodge a ferrite inhibition test; pick tin if the master is dirty and the part is a short-run magnet with loose tolerances.

Moisture in a bag of ferrite is a tin-cure variable and a void source in both systems. Keep powder dry. Do not oven-burn a polymer coating off a treated ferrite unless the compounder says so.

Spec the magnet, then pick the process

Write function and constraints:

• Pull or holding force on a named steel at a named air gap and part thickness, or a flux-density method on a named coupon if the customer actually has a Gauss meter procedure. Do not put a forum Gauss number on a drawing with no geometry.
• Flexible magnet versus flux path (hard ferrite vs soft ferrite).
• Process: platinum RTV prototype, tin RTV only if justified, magnetic HCR, or magnetic LSR.
• Homogeneity: density range, no sludge line, pull coupons from opposite ends of the shot.
• Mechanicals on the filled compound (ASTM D412 / D624). Unfilled twins are fiction.
• Tooling: abrasive compound, steel for production, not a printed cavity for a magnetic LSR.
• Alignment field: yes/no.
• Contact rules: magnetic toys and sensors are not automatically food-grade. FDA 21 CFR 177.2600 and ISO 10993 are finished-formulation problems. Ferrite dust on a food mold is a contamination failure.

Refuse:

• “Make it magnetic” with no pull, no gap, no thickness
• A customer-supplied ferrite of unknown coating dumped into a platinum 1:1 kit
• Copying an HCR loading into an RTV cup
• Claiming EMI shielding because the part is magnetic (that is a conductivity and percolation problem; see the conductive-filler article)

If the part only needs to locate on a fixture, sometimes a bonded steel insert or a discrete magnet overmolded in unfilled silicone is cheaper and stronger than a barely loaded RTV. Magnetic rubber is for when the rubber itself must be the magnet.

FAQ

How much ferrite should I add to platinum RTV?

Do not invent a percent. Loading is a compounder’s trade between pull and a mix you can still wet and degas. RTV cups accept less ferrite than a two-roll HCR mill. Specify holding force or a coupon flux method on the geometry, then use a formulated grade or a tested let-down.

Why did my magnetic RTV pour cure sticky on the bottom?

The sludge layer is the highest ferrite and the most contamination. Some ferrite coatings and dirty iron oxides interfere with platinum addition cure. Test the actual powder in the actual base, including a settled cup. Tin-cure is not an automatic fix; it only changes the catalyst problem and still shrinks with time.

Can I stop ferrite from settling by mixing longer?

Longer mixing wets better and then puts more air in. Once you stop, dense ferrite falls through a thin RTV. Use a thixotropic base, mix then pour without a long rest, and check density or pull at both ends of the part. Vacuum after the powder is wet, not on a dry dump.

Is magnetic LSR just RTV with more powder?

No. LSR needs an injectable rheology, meter-mix, and a tool that can take an abrasive, high-density compound. Weld lines can be ferrite-poor (dead magnetic zones). Production flexible magnets are often HCR milled grades. Copying a mill loading into a cup is how you get a paste you cannot pour.

Does a stronger mix mean I can skip magnet orientation in the tool?

Not if the drawing needs a pole face. Hard-ferrite compounds get a large share of their useful pull from alignment in a field during cure. A random RTV pour is weaker for the same powder. If orientation is required, that is a fixture and a production process, not a longer stir.