A Magnet Is Not a Metal Test: What Buyers Get Wrong About Non-Magnetic Alloys

The question sounds practical, and that is what makes it dangerous. Ask which metals do not stick to a magnet and you get an answer shaped like a list — aluminium, copper, brass, bronze — and that list looks usable right up to the moment a shipment arrives and the magnet grabs something it should not have. There is no such list, because metals do not divide into magnetic and non-magnetic camps. Magnetism is a behaviour that belongs to a base element and to the phase that element happens to be in when the magnet comes near, which means one line on a purchase order can cover material that answers a magnet and material that ignores it. On the buying side, a wrong assumption is paid for in returns, rework and schedule slips rather than in a test report, which is why a magnet is the most overrated tool in the room. It is not that the tool lies; it is that it is being asked a question it cannot answer.

A Magnet Cannot Sort Metals Into Two Camps

The question sounds practical, and that is what makes it dangerous. Ask which metals do not stick to a magnet and you get an answer shaped like a list — aluminium, copper, brass, bronze — and that list looks usable right up to the moment a shipment arrives and the magnet grabs something it should not have. There is no such list, because metals do not divide into magnetic and non-magnetic camps. Magnetism is a behaviour that belongs to a base element and to the phase that element happens to be in when the magnet comes near, which means one line on a purchase order can cover material that answers a magnet and material that ignores it. On the buying side, a wrong assumption is paid for in returns, rework and schedule slips rather than in a test report, which is why a magnet is the most overrated tool in the room. It is not that the tool lies; it is that it is being asked a question it cannot answer.

Ferromagnetism is not a property of metals in general; it belongs to a handful of base elements, chiefly iron, cobalt and nickel, whose atomic moments can be aligned by an applied field. Build an alloy on aluminium or copper and that mechanism is simply absent, so the questions worth asking change entirely. The axes that actually separate one copper alloy from another are conductivity, corrosion behaviour, machinability and wear resistance, and they separate them sharply. Copper leads them at 101% IACS, which is why it goes into current-carrying parts. Brass C36000 is chosen for high-speed CNC machining and cost efficiency. Bronze C93200 is chosen for anti-friction and wear resistance. Choose wrongly and the consequences are specific: connectors that overheat because the conductivity was never there, marine fittings that seize because the corrosion behaviour did not suit saltwater. None of those four axes appears in a magnet's reaction, which is the first place a magnetic screening habit goes wrong.

Strip the question down and what remains is a three-part identification: what the base element is, which alloy system it belongs to, and what state the material is in as delivered. All three are answerable before anything is ordered, all three are already tracked by the mill and the distributor, and none of them needs a magnet. What a magnet reports is a fourth thing — whether the piece in front of you currently contains a ferromagnetic phase in a quantity and distribution a small field can detect. That observation is real and occasionally useful, but it sits downstream of the decisions that settle whether the material suits the job. Everything below takes that order seriously: the non-ferrous families, the steel that breaks the magnetic rule, and the screening sequence that should replace the magnet at the desk.

Why Aluminium, Copper, Brass and Bronze All Ignore a Magnet — and Why That Proves Nothing

The closest thing to the list people want is the non-ferrous family, and as a family it is genuinely non-magnetic. Aluminium, copper, brass, bronze and the beryllium coppers catalogued beside them all sit on bases that are not ferromagnetic, so a magnet cannot separate one from another. Copper alloys show how much that single reading hides: the differences that matter sit in alloy data, not in magnetic response. A manganese bronze such as C86300 earns its place on exceptional strength and good wearing properties while machining only fairly, and a phosphor bronze such as C90700 offers high strength and corrosion resistance for heavy-duty gears and bearings. These are hard to tell apart by eye, which is why designers and buyers confuse them. Put a magnet on both and you learn one fact, which is nothing.

Aluminium makes the same argument differently, because its selection matrix is wide enough to show how much real variation lives inside one non-magnetic material. Sheet stock typically runs from 0.2mm to 6.0mm in thickness, and a buyer choosing it is choosing an alloy and a temper at once: 1050 and 3003 where formability and corrosion resistance lead, 5052 where strength and weldability matter, 6061 where machining and structural performance dominate. Temper — H14, H32, T6 — describes the hardness and mechanical treatment the material has received, and it is not a formality. The wrong temper shows up as cracking during bending, excessive springback, poor surface finish, or fabrication that is harder than it needs to be. Run a magnet across that whole matrix and the reaction is identical everywhere, which means the observation is blind to the decision that matters most in the catalogue.

Inside the copper family the real fork is brass against bronze, and it is decided by what the part has to survive rather than by which metal is better. Both are copper alloys and they behave differently the moment the cutter touches them. Brass is the choice for high-volume precision parts with thin walls or fine cosmetic finishes, where easier cutting becomes faster cycles and lower tool wear. Bronze earns its place in bearings, bushings and load-bearing or corrosion-prone components, where wear resistance and durability carry the decision. Machinability, corrosion behaviour, conductivity and cost move in different directions across the two, so the problem is one of matching material to job rather than ranking. Swap one for the other and nothing in the magnetic reading changes, which is precisely why the reading cannot be trusted with the choice.

The Steel That Doesn't Stick: an 18/10 Pan Breaks the Rule

Take a 12-inch three-ply skillet off a shelf and the assumption collapses in your hand. The pan is fully bonded — 18/10 stainless as the exterior, a pure aluminium core, 18/10 stainless again as the cooking surface — and it was tested for a week across gas, induction and electric, scoring 5 out of 5 for performance and build quality. 18/10 is the austenitic grade, and the phase structure that makes a stainless steel tough and corrosion resistant is also the one that largely declines to align with a small magnetic field, so the reading at the cooking surface is not what the steel rule predicts. The same pan is sold as induction compatible, which is a magnetic interaction by definition. Both statements are true at once, and together they destroy the shortcut: what a magnet reports depends on which layer and which phase the field reaches, and the answer can vary within a single object.

Temper keeps proving the same point: an alloy name is only half a material's identity, and the other half is the state it arrives in — how hard it is and how it was worked to get there. Get that state wrong and the shop pays: a blank that splits at the bend line, a part that springs back past what the tooling allows, a surface that will not finish. Steel follows the same logic, so a magnet reading is not a property of the grade on the certificate but of the condition of the piece in front of you. A grade that leaves the mill in a soft austenitic condition may respond weakly or not at all, while the same grade after heavy cold forming can shift the reading. Same designation, same paperwork, different answer — a buyer who treated the reading as an identity test has confused a processing variable with a material variable.

So the honest description of what a magnet measures is narrow: the quantity and distribution of ferromagnetic phase in the piece at the moment of testing. That is a genuine measurement, and it can occasionally rule something out, but it is not a grade, not a purity figure and not a quality rating — and it does not become one by being convenient. This is where purchasing habits do their damage. A buyer who has accepted the idea that steel is magnetic as a rule has also, usually without noticing, accepted that a non-magnetic steel is fake, that a magnetic non-steel is contaminated, and that the strength of the pull tracks how good the material is. All three conclusions are wrong in the same way: they convert an observation about phase into a claim about identity.

Three Ways a Magnet Misleads a Buyer at the Desk

The first misread is the quiet one. Two copper alloys look alike, both ignore the magnet, so they get treated as interchangeable and price breaks the tie. That is the wrong tiebreaker, because brass and bronze diverge the moment the part goes to work. Machinability announces itself immediately — tool wear, cycle time, surface finish, scrap rate — and it decides the economics of any high-volume job that has to hold thin wall sections or come out cosmetically clean, which is the work brass is bought for. Wear resistance and durability announce themselves later, usually at the failure, in parts that carry a load or face saltwater, which is bronze's territory. A decision settled by unit price and confirmed by a magnet has optimised the cheapest variable and ignored the one that generates returns.

A distributor's material selector is more instructive than it looks. When the dropdown reads alloy steel, aluminium, beryllium copper, brass, bronze, carbon steel, the platform is sorting inventory by base material before it asks anything else, and that is the order a buyer should use. Lock the base element first, because that is what decides whether ferromagnetism is even possible and which family of properties you are shopping in. Fix the form and the state second — sheet or plate, bar or coil, which temper, which thickness range — because those decide whether the material will form, machine or weld in your process. Only then compare applications and cost against the intended service condition. Nothing in that sequence asks for a magnet; if it comes first in yours, the sequence has been inverted.

The third misread is about scale, and it should retire the magnet from the front of the process. Non-ferrous material is not the exception in the supply chain; it is a mainstream industrial flow. Among aluminium producers, Alcoa is one of the largest globally and runs a vertically integrated operation from bauxite onward, with a 40-acre campus at New Kensington. That business announced a 10-year, $1.1 billion contract with Pratt & Whitney covering the first advanced aluminium alloy fan blade technology for its engines, and it spent $2.85 billion to acquire a UK jet-engine component maker. Industries with the tightest tolerances and the harshest failure consequences are built on a metal a magnet ignores. If aluminium needed magnetic proof of legitimacy, none of that investment makes sense.

The Rule Worth Keeping: a Magnet Can Exclude, It Cannot Determine

Here is the rule worth keeping. Name the base element first, because that is what makes ferromagnetism possible at all; if the base is aluminium, copper or a copper-alloy system, the magnet question is settled before the material is ordered. Name the state the material is supplied in second — temper, condition, form — because that is where the behaviour that affects your process lives, and where paperwork and a physical test can legitimately disagree. Match the application and the cost against the service condition third, and only third. The catalogue you order from already works this way: raw materials are grouped as metals, plastics and other stock, categories follow material family and product form, and the documents that settle disputes sit in a quality section rather than a tool kit. Nowhere in those three steps does a magnet earn a place.

The boundary matters as much as the rule. A magnet is a reasonable exclusion tool: if you expected ferromagnetic behaviour and get none, or the reverse, it is worth pausing before the material goes into a machine. What it cannot do is produce a conclusion about a grade or a lot, because that conclusion requires documentation — the test reports and quality paperwork that suppliers of certified material issue with the shipment — rather than a field observation about attraction. The leftover problem keeps the rule honest: the same specification, processed differently, can answer a magnet differently, so even a documented grade is not guaranteed to give the reading you have learned to expect. Use the magnet to raise a question, and the paperwork to answer it. When the two disagree, the paperwork wins.

Which metals do not stick to a magnet? The list people are looking for does not exist, and the reason is not missing data — it is that the question sorts material by a behaviour instead of by an identity. What does exist is a sequence: base element, then state, then application. Run it and the magnet becomes what it should have been all along, an optional spot check at the end rather than a verdict at the start. The question worth carrying into the next order is narrower: for a given grade and temper, how far does the magnetic response actually move with processing, and is that movement large enough to matter for the part I am about to make?